Framework for Integrating Requirements Engineering and DevOps Practices in Robotic Process Automation with a Focus on Optimizing Human–Computer Interaction

Abstract

This study introduces FRIDA, a novel framework that integrates Requirements Engineering (RE) and DevOps practices into Robotic Process Automation (RPA), with a focus on optimizing human–computer interaction (HCI). The framework was designed using the PICO (Population, Intervention, Comparison, and Outcome) methodology and evaluated through a case study in an automation-driven department. Key results include an 83% reduction in processing time, an 81.25% decrease in error rates, and an 80% reduction in manual tasks, alongside improved compliance and scalability. The integration of RE and DevOps ensures a structured approach to requirement management and continuous delivery, while the emphasis on HCI enhances user-friendliness and adoption. FRIDA represents a significant advancement in RPA development, offering a robust, user-centric solution for optimizing automated processes. Future research should explore its application across diverse industries and the integration of advanced technologies like AI and machine learning to further enhance RPA capabilities.

1. Introduction

Robotic Process Automation (RPA) has emerged as a transformative technology in the realm of digital transformation, enabling organizations to automate repetitive, high-volume tasks traditionally performed by humans. As adoption of this technology grows, the integration of best practices from Requirements Engineering (RE) and DevOps has become crucial for developing efficient, high-quality RPA systems. Requirements Engineering focuses on identifying, analyzing, and specifying stakeholder needs, ensuring that systems meet both functional and non-functional requirements. DevOps, on the other hand, bridges the gap between development and operations, emphasizing automation, continuous collaboration, and agile software delivery to streamline the construction, testing, and deployment of solutions [1].

RPA has rapidly evolved as a powerful tool for automating rule-based tasks, reducing operational costs, and improving efficiency [2]. However, despite its growing adoption, significant challenges persist in the development and implementation of RPA systems. A primary issue is the lack of a structured approach to Requirements Engineering, which often results in poorly defined automation goals, misaligned stakeholder expectations, and suboptimal system performance [3]. Additionally, the integration of DevOps practices into RPA projects remains underdeveloped, leading to fragmented development cycles, delayed deployments, and difficulties in maintaining and scaling automated processes [4]. Furthermore, the role of human–computer interaction (HCI) in RPA is frequently overlooked, resulting in systems that are not user-friendly, difficult to monitor, and prone to low user adoption rates [5]. These challenges are exacerbated by the increasing complexity of business processes, the need for regulatory compliance, and the demand for faster, more agile automation solutions [6].

Recent studies highlight that while RPA adoption is on the rise, many organizations struggle with high failure rates due to inadequate planning, insufficient testing, and poor alignment between technical capabilities and business needs [7]. For instance, research indicates that 30–50% of initial RPA projects fail to meet expectations, often due to a lack of clear requirements and an insufficient focus on user experience [8]. Moreover, the emergence of Low-Code/No-Code platforms and Open-Source RPA tools has introduced new opportunities for democratizing automation but also poses challenges in terms of governance, security, and integration with existing systems [9]. The absence of a framework that integrates RE, DevOps, and HCI practices exacerbates these issues, leaving organizations without a clear roadmap to optimize their RPA initiatives [10]. Recent research underscores the need for a unified approach that addresses both technical and operational aspects while prioritizing user-centric design to ensure the sustainability and scalability of RPA solutions [11].

This study seeks to address these gaps by proposing a comprehensive framework that integrates RE and DevOps practices into the RPA lifecycle, with a strong emphasis on enhancing HCI. By doing so, it aims to provide a structured, user-focused methodology that improves the efficiency, adaptability, and usability of RPA systems, ultimately contributing to higher success rates and greater business value [12].

The integration of RE and DevOps in RPA development can significantly optimize the project lifecycle, ensuring that requirements are properly identified and addressed throughout all phases of development and operation. For example, RE practices can be employed to gather clear and complete requirements before automation development begins, while DevOps can ensure continuous and agile updates and deployments without disrupting automated processes [13]. This combined approach not only improves the efficiency of the automation process but also creates a more robust system aligned with the real needs of users.

Simultaneously, human–computer interaction (HCI) plays a pivotal role in ensuring that RPA solutions are intuitive and effective for end-users. Focused on user experience, HCI aims to provide seamless and efficient interaction between humans and machines, which is particularly relevant in automation systems that directly impact user performance and satisfaction. Examples of effective HCI application in RPA include user-friendly control interfaces for automated processes that are both easy to use and powerful enough to enable customization and monitoring by users [14].

The motivation for this study stems from the prevalence of manual processes across various industries and the urgent need to integrate practices such as RE and DevOps to ensure that RPA systems are developed in a more efficient and responsive manner. By integrating these practices, organizations can provide a more intuitive and optimized user experience while significantly improving operational efficiency. Additionally, the rise of Low-Code automation and Open-Source tools presents new opportunities and challenges in building RPA solutions that meet diverse and complex requirements [14].

This study aims to develop a framework that integrates RE and DevOps practices into the lifecycle of RPA projects, with a particular focus on improving HCI. The framework seeks to enhance the efficiency of RPA development while providing a smoother and more optimized experience to end-users. Its practical application will be demonstrated through a real-world case study, highlighting benefits such as time savings and cost reduction in automation projects.

The structure of this article is as follows: Section 2 provides a comprehensive review of the state of the art, exploring the concepts of RPA, RE, DevOps, and HCI, as well as their interrelationships. Section 3 outlines the adopted methodology, detailing the process of hypothesis formulation and the PICO approach. Section 4 proposes an innovative framework and its benefits. Section 5 illustrates the practical application of the framework through a case study. Section 6 analyses the results obtained, while Section 7 concludes the article, summarizing the key contributions and suggesting directions for future research.

2. Background

Robotic Process Automation (RPA) is an emerging technology that facilitates the automation of repetitive, rule-based tasks traditionally performed by humans. RPA employs “bots” or intelligent software to mimic user actions on existing system interfaces, thereby enhancing data processing efficiency and streamlining business operations. This form of automation is particularly effective in environments where processes adhere to well-defined and structured workflows, enabling tasks such as data entry, information validation, and transaction processing to be executed seamlessly and without error [15].

RPA has been deployed across various industries, including finance, human resources, and customer support. In the financial sector, for instance, RPA bots are utilized to automate invoice processing, bank reconciliations, and financial report analysis. Within human resources, RPA streamlines recruitment processes, such as CV screening and interview scheduling. In customer support, RPA bots assist in responding to frequently asked questions and automating service tasks, thereby enhancing both efficiency and service quality. The advantages of RPA are manifold, including heightened productivity, reduced human error, improved operational efficiency, and the potential to free employees to engage in higher-value activities [16].

Requirements Engineering (RE) is the discipline concerned with identifying, analyzing, documenting, and managing the requirements necessary for software system development. Its significance in software engineering is paramount, as the accurate definition of requirements ensures that the system meets stakeholder needs, thereby mitigating rework and implementation failures. In the context of RPA, RE is crucial, as the manual processes slated for automation must be thoroughly understood before they can be effectively replicated by bots [17].

The RE processes applicable to RPA projects encompass requirement gathering techniques such as stakeholder interviews, process analysis, and the identification of inefficiencies in existing operations. Moreover, the analysis of dynamic requirements is a key aspect, as business needs and conditions may evolve throughout the project lifecycle. RE is also instrumental in mapping the requirements of manual processes to be automated, ensuring that bot operations are effective, secure, and aligned with user expectations [18].

DevOps is a methodology that fosters the integration of development and operations teams to enhance collaboration, accelerate delivery cycles, and uphold software quality. A fundamental principle of DevOps is the automation of processes such as testing, integration, and continuous deployment, all of which contribute to more agile and efficient software development. The incorporation of DevOps practices into RPA projects enables continuous integration (CI), ensuring that RPA bots can be updated and deployed seamlessly without disrupting routine operations [19,20].

Human–computer interaction (HCI) pertains to the design and study of interactions between humans and computer systems, with a strong emphasis on user experience. HCI is critical in the development of interactive systems, such as RPA solutions, as effective interaction contributes to usability, efficiency, and overall user satisfaction. In RPA projects, the interaction between humans and bots must be carefully designed to ensure intuitive and effective user engagement [21].

HCI directly influences the user experience in RPA systems, where well-designed interactions facilitate smooth and seamless operations, whereas poor interactions can lead to frustration and system rejection. Notable examples of interaction design in RPA include real-time feedback mechanisms, user-friendly interfaces, and continuous user experience monitoring to ensure that RPA bots meet user expectations and deliver substantial value in daily workflows [22].

The adoption of emerging technologies such as RPA necessitates critical engagement and the thorough evaluation of available solutions. Organizations aiming to automate operational processes must consider not only tangible benefits such as cost reduction and efficiency gains but also intangible factors such as usability, integration with legacy systems, and employee acceptance.

Critical engagement involves recognizing that not all cutting-edge solutions are suitable for every business context. RPA implementation should be preceded by an in-depth analysis of existing processes, identifying bottlenecks and opportunities for optimization. Furthermore, strategic alignment between technology and organizational objectives is imperative to ensure that automation delivers tangible value.

Another key consideration is the human impact of automation. Resistance to RPA adoption may emerge due to concerns over job displacement, necessitating proactive change management strategies. Investment in training and the redefinition of organizational roles are essential to facilitate a smooth transition. Consequently, implementing RPA not only enhances operational efficiency but also strengthens employees’ roles, allowing them to focus on higher-value tasks.

Integrating Requirements Engineering, DevOps, and RPA presents multiple challenges that must be addressed to ensure successful automation projects [23].

To guarantee the efficacy of RPA-driven process automation, a structured technical approach focused on process readability must be adopted. From a technical perspective, requirements analysis should be conducted with utmost precision, ensuring that business rules are well-defined and comprehended before implementation. This approach mitigates risks associated with erroneous or inconsistent process automation, ensuring that bots function efficiently.

Integrating RPA with existing systems presents a technical challenge, requiring compatibility with APIs, databases, and communication protocols. Selecting RPA platforms that support native integration with enterprise systems such as ERPs and CRMs simplifies deployment and reduces development time. Additionally, information security must be prioritized to ensure compliance with regulatory guidelines and safeguard sensitive data.

The readability of automated processes is another critical factor. Comprehensive documentation, including flow diagrams and detailed process descriptions, facilitates the comprehension and maintenance of bots. Practices such as script versioning and the implementation of logs to monitor execution are indispensable for ensuring the transparency and auditability of automated operations.

Lastly, the continuous monitoring of bots is essential to detect potential failures and implement necessary adjustments promptly. The integration of techniques with rigorous requirements, secure system integration, and meticulous documentation practices ensures that RPA is deployed in a robust, comprehensible, and efficient manner, thereby maximizing its positive impact on organizations.

• Challenges in Requirements Engineering: The complexity of defining requirements for RPA processes can be a significant impediment, particularly as these requirements may evolve throughout the project’s lifecycle. Ambiguities or a lack of detail in requirement specifications can adversely affect RPA implementation. Additionally, the dynamic nature of requirements necessitates a flexible and adaptive approach to Requirements Engineering [23].
• Challenges of DevOps in RPA Projects: Adapting traditional software development methodologies to accommodate the specific demands of RPA presents further challenges. The automation of testing and monitoring, as well as the continuous deployment of RPA bots, necessitates specialized tools and methodologies, as well as the capability to manage and oversee bot updates in real time. Continuous collaboration between development and operations teams is essential in order to address the complexity and frequency of updates [23].
• Challenges in human–computer interaction: The interaction between humans and bots may be hindered if RPA systems are not optimally designed, leading to issues such as poor usability, lack of transparency, and diminished trust in automated bots. Addressing these challenges requires meticulous interface design and the implementation of HCI best practices to ensure an optimal user experience, thereby fostering user acceptance and the effective utilization of RPA bots [23].

3. Methodology

3.1. Method

The PICO approach (Population, Intervention, Comparison, Outcome) has proven to be an essential tool in scientific research due to its ability to structure and organize complex questions in a clear and systematic manner, as shown in

Table 1. PICO model structure.

P	(Population) Group of interest in the study.
I	(Intervention) Exposure or factor analyzed.
C	(Comparison) Comparison with another intervention or control group.
O	(Outcome) Expected or measured result of the study.

. When applied rigorously, the PICO methodology allows for a detailed and objective analysis of the key elements of a study, avoiding the subjective biases that are often present in other approaches, such as narrative reviews [24]. The PICO methodology has not only stood out in fields such as healthcare and medicine but has also demonstrated its value in areas such as software engineering, automation processes, and agile development practices, as evidenced by recent research (to include updated references). It facilitates the formulation of precise questions and the selection of relevant studies, ensuring that data are analyzed consistently and comparably [25].

In this study, the PICO methodology is applied to structure the analysis of the integration of Requirements Engineering (RE) practices and DevOps in the development of Robotic Process Automation (RPA) systems, with a particular focus on optimizing human–computer interaction. The adaptation of this methodology to the context of Requirements Engineering, DevOps, and RPA provides a systematic approach to understanding how the integration of these components impacts the lifecycle of an automation project, with special attention to user experience and the performance of the automation system [26,27].

In the present work, the “population” refers to organizations that have already implemented or are planning to adopt RPA solutions via the integration of RE and DevOps practices, with a particular focus on improving human interaction with the automated system. Defining this population precisely is crucial to ensure that the results are relevant and applicable to industrial and business contexts that face similar challenges related to process automation and the development of user-centric systems. By examining companies that apply advanced RE and DevOps practices to create more effective, user-centric RPA solutions, a deeper understanding of development processes and their impacts on organizational performance and user experience can be gained.

The “intervention” in this study involves the implementation of combined RE and DevOps practices in the development of RPA systems, with the aim of optimizing human–computer interaction and improving the efficiency of automated processes. This intervention is based on growing evidence that integrating these practices can not only increase the efficiency and quality of RPA system development but also enhance the user experience by creating intuitive and highly functional systems.

The “comparison” in this study will be conducted between the manual execution of a process within a specific department of a company and the implementation of the framework developed in this work, which incorporates integrated RE and DevOps practices into the automation of RPA projects for the same process. This analysis will evaluate the differences before and after the framework’s application, taking into account the gains and benefits achieved through automation. The comparison aims to identify improvements in the process lifecycle efficiency and the advantages realized following the framework’s implementation.

The “outcome” expected from this study is to demonstrate that integrating RE and DevOps practices into RPA projects contributes to more efficient development, enhancing both human–computer interaction and operational outcomes. By applying the PICO methodology, it will be possible to formulate solid hypotheses and conduct a grounded analysis, based on concrete data, that will validate the effectiveness of this approach in the context of process automation and systems development.

The methodology adopted for this study involved the thorough analysis of relevant sources and the application of the PICO approach for the selection of articles, with strict inclusion and exclusion criteria. The selection of articles was based on a detailed bibliographic search and the careful screening of titles and abstracts, prioritizing studies that address the integration of RPA, RE, DevOps, and human–computer interaction. Articles that did not meet the relevance criteria were discarded, resulting in a final selection of significant studies for the research. Data collection was carried out from reputable scientific databases, such as the “B-on” platform, ensuring that the selected sources were reliable and up to date [28].

The application of the PICO methodology provides a robust framework for organizing the analysis of the collected data, demonstrating the clear connection between the hypotheses formulated and the evidence found in the literature [28]. Thus, the approach facilitates the interpretation of results and the development of an effective framework for integrating RE and DevOps in RPA projects, with a particular focus on optimizing human interaction with automated systems.

The inclusion criteria adopted for this study were as follows:

• Articles that directly discuss the implementation or evaluation of RPA and RE and DevOps practices in process automation, focusing on optimizing human–computer interaction.
• Studies conducted between 2010 and 2025, ensuring the relevance and timeliness of the practices and technologies discussed.
• Articles published in peer-reviewed scientific journals and available in full text.

The following types of publications were excluded:

• Articles that do not directly address the implementation of RPA, RE, and DevOps or human–computer interaction, or that focus on areas unrelated to process automation.
• Works that do not significantly discuss aspects of optimizing human–computer interaction in RPA systems or that do not address integration with RE and/or DevOps practices.
• Publications that are inaccessible in full text or with restricted access.
• Studies that are not based on empirical methodologies or that do not present clear and consistent data on the impacts of integrating RPA, RE, DevOps, and HCI on the development and operation of automated systems.

This rigorous selection process ensured that the analysis was based on high-quality sources directly relevant to the study of integrating RPA, RE, DevOps, and HCI in the development of automated systems, with a focus on optimizing the user experience and improving operational efficiency.

The central research question and the hypotheses that guided this study were formulated as follows:

Central research question (CRQ)

CRQ. 

Information obtained from the user’s participation in the human–computer interaction (HCI) process can be readily adapted to functional requirements during the development of Robotic Process Automation (RPA) systems.

Hypotheses (H)

H1. 

The integration of RE and DevOps practices improves operational efficiency in the development of RPA systems, reducing development cycle time and enhancing the continuous delivery of automation solutions.

H2. 

The combination of RE and DevOps, with a focus on human–computer interaction, results in more intuitive and user-friendly RPA systems, improving user experience and the adoption of automated solutions.

This study aims to provide an in-depth understanding of how integrating RE and DevOps practices can optimize RPA system development, improve human–computer interaction, and increase operational efficiency within organizations. By adopting these practices, the study seeks to demonstrate the benefits in terms of cost reduction, software quality improvement, and greater end-user satisfaction.

The two hypotheses presented support the idea that combining RE and DevOps practices in RPA system development brings operational benefits and enhances user experience. The first hypothesis (H1) suggests that integrating RE and DevOps into the RPA development lifecycle leads to more agile and efficient processes, reducing development time and ensuring the continuous delivery of high-quality automation. This improvement in the development process can optimize operational efficiency, lowering costs and enhancing system resilience.

The second hypothesis (H2) proposes that focusing on human–computer interaction, alongside the application of RE and DevOps practices, creates more user-friendly and intuitive RPA solutions. This improves user adoption and contributes to a better user experience, a critical factor in the success of automated solutions. Furthermore, by ensuring that the systems better meet user needs, this model can optimize human resources, allowing professionals to focus on higher-value strategic tasks while enhancing operational efficiency and overall satisfaction.

To address the central research question of this study, the researchers utilized the online scientific library provided by the Foundation for Science and Technology, focusing on four specific groups (Group 1, Group 2, Group 3, and Group 4), as detailed in

Table 2. Groups searched through “B-on”.

Group 1	Group 2	Group 3	Group 4
“RPA” OR “Robotic Process Automation” OR “Intelligent Process Automation” OR “Digital Process Automation” OR “Business Workflow Automation”	“Requirements Engineering” OR “Requirement Analysis” OR “System Requirements” OR “Requirement Specification” OR “Requirement Validation” OR “Requirement Management” OR “Functional Requirements” OR “RE”	“DevOps Practices” OR “Continuous Integration” OR “Continuous Deployment” OR “Infrastructure as Code” OR “Agile Development” OR “DevOps Automation” OR “CI/CD Pipeline” OR “DevOps”	“Human-Computer Interaction” OR “User Experience” OR “Interface Design” OR “Usability Testing” OR “Interaction Design” OR “User-Centered Design” OR “Cognitive Load” OR “HCI”

.

The investigation was conducted using the “B-on” platform, employing the OR operator to connect either the title, keywords (KWs), or abstract (AB) within the three designated groups. Organizing keywords into separate sets, grouped according to their theme and composed of synonyms, is an extremely effective approach to optimizing the search for articles on scientific platforms and databases. This structuring allows each set to represent a specific concept, ensuring that the research covers all possible linguistic and terminological variations of the same term, including different synonyms, plurals and morphological variations. When these sets of keywords are combined in the search using the Boolean operator “AND”, it ensures that only articles that address all the fundamental concepts are retrieved, providing highly relevant and accurate results. This significantly reduces the number of irrelevant documents, as it avoids retrieving articles that only mention part of the topic without addressing the entire subject of interest. Furthermore, this methodology increases the effectiveness of research by improving the accuracy of results, allowing researchers to find studies that truly match their objective more quickly and assertively. Thus, instead of obtaining thousands of scattered and often disconnected articles, the researcher accesses a more restricted but highly qualified set, optimizing their time and facilitating the analysis of the available data.

Subsequently, filters were applied to the collection of publications gathered during the research process, and the findings, in terms of the total number of publications, are outlined in

Table 3. Publications obtained through B-on, after the application of some filters.

	Set 1	Set 2	Set 3
Initial result:	806	18	51
1—Restrict to peer-reviewed	378	18	14
2—From 2010 to 2025	318	18	13
3—Language: English	299	13	6
4—Restrict to full text	274	7	3

.

After applying the filters, the titles, keywords, and abstracts of each paper were reviewed to select those most relevant to the research topic. Initially, 875 articles were retrieved. After the filters were applied, 284 remained, of which only 15 were directly related to the study’s focus.

Figure 1 presents a flowchart describing the systematic literature search and selection process carried out in this study. The process begins with the identification of search terms, which are used to formulate the search string, resulting in an initial total of 875 articles. Then, filtering criteria are applied to refine the results:

• Restriction to peer-reviewed articles, reducing the number to 410 articles.
• The limitation of the publication period, considering only articles published between 2010 and 2025, which results in 349 articles.
• Selection by language, keeping only articles in English, which reduces the total to 318 articles.
• Restriction to articles with full text available, leaving 284 articles for analysis.

Figure 1. Flow diagram of literature search and respective screening.

Figure 1. Flow diagram of literature search and respective screening.

After filtering, the 284 articles are subjected to a relevance analysis, where each publication is evaluated to determine whether it is directly related to the study topic. Articles that do not meet this criterion are excluded, while those that are considered relevant are selected for detailed analysis. At the end of the process, 15 articles are identified as directly related and included for analysis in the study.

This flowchart illustrates in a clear and structured way the methodological rigor applied in the literature review, ensuring the selection of relevant and high-quality sources to support the research.

3.2. Articles Synthesis and Analysis

This section offers a comprehensive overview and analysis of the articles most relevant to the subject under investigation.

Table 4. Identified articles and the respective themes of the articles found.

	Themes of the Articles	HCI	RPA	DevOps	RE
Articles
[29]		X			X
[30]			X
[31]			X
[32]		X
[33]			X
[34]				X	X
[35]			X
[36]			X		X
[37]				X
[38]				X
[39]				X
[40]			X
[41]		X			X
[42]			X
[43]		X		X
% Themes p/articles		27%	47%	33%	27%

, presented below, displays the 15 chosen articles along with the models examined in each. The table was developed to categorize the contributions of each study and was meticulously created following an in-depth search of academic databases.

The selected articles were carefully reviewed and analyzed to pinpoint the main themes and methodologies addressed. These themes were then arranged in a table, where each column represents a specific theme, and the rows list the articles, annotated according to the models discussed.

For the selection of the articles analyzed in this study, strict criteria were considered that guarantee the relevance and quality of the information used. The main inclusion criteria included publications in high-impact scientific journals, articles published in recent years to ensure current information, studies with a clear and well-defined methodology, and research that directly addressed the topic under study. Furthermore, articles without peer review, publications in languages not accessible to researchers, and studies that did not present sufficient empirical data for analysis were excluded.

After selecting the articles, detailed analysis was carried out to identify the main themes addressed and their correlation. It was observed that the selected studies cover fundamental aspects of the problem in question, being divided into distinct but interrelated categories. The analysis of the correlation between the themes revealed important patterns, showing that certain factors directly influence others, forming a complex network of interrelationships. This approach not only allows the mapping of trends in the area, but also the identification of gaps in the literature, suggesting directions for future research.

The use of this analysis method is justified by its ability to structure knowledge in a systematic and in-depth manner. By categorizing and correlating themes, it is possible to obtain a broad and integrated view of the topic studied, facilitating the identification of patterns and the formulation of hypotheses for future investigations. In this way, this approach contributes significantly to the construction of a solid and well-founded theoretical framework.

3.3. Discussion of Literature Review Results

In this section, the results of the articles reviewed in the previous section are synthesized, highlighting the main contributions, common themes, and potential gaps in the literature. The themes—human–computer interaction (HCI), Robotic Process Automation (RPA), DevOps, and Requirements Engineering (RE)—are analyzed for their prevalence in the selected articles, and the relationships between them are discussed.

As shown in Table 3, each article discusses one or more of the four key themes. The distribution of these themes across the articles is as follows:

• Human–computer interaction (HCI): featured in 27% of the article, this theme is primarily explored in the context of improving user experience and optimizing human interaction with automated systems.
• Robotic process automation (RPA): this most prevalent theme, appearing in 47% of the articles.
• DevOps: discussed in 33% of the articles, this theme focuses on the integration of automation into the software development lifecycle.
• Requirements Engineering (RE): identified in 27% of the articles, this theme explores the need for detailed Requirements Engineering in the development of automated systems.
• Synergies between the themes: The intersection of RPA with HCI and DevOps stands out as a key area of focus. RPA systems benefit significantly from HCI principles, which help in designing more user-friendly interfaces, thereby improving system adoption and efficiency. Furthermore, the integration of DevOps practices in RPA development enables continuous integration and deployment, ensuring that automation solutions are scalable and adaptable.
• Challenges in RE: Requirements Engineering (RE) is frequently discussed in relation to the early stages of system development, especially in RPA projects.

Despite the progress in each theme, several gaps merit further investigation:

• Lack of unified frameworks: While the individual themes are well explored, there is a lack of a comprehensive approach that integrates HCI, RPA, DevOps, and RE in a cohesive manner. The development of such frameworks could provide more practical guidance for the implementation of automation systems.
• Industry-specific applications: many of the studies focus on general applications of RPA, with limited exploration of sector-specific adaptations.

4. The Proposed Method

4.1. Proposal for a Framework (FRIDA)

This chapter presents an innovative framework, named FRIDA, designed to integrate Requirements Engineering (RE) practices and DevOps methodologies into the development of Robotic Process Automation (RPA) systems, with a particular emphasis on optimizing human–computer interaction (HCI). The objective of this model is to enhance operational efficiency, streamline automation development processes, and improve user experience by creating more intuitive and responsive systems. FRIDA—the Framework for Requirements Engineering and DevOps Integration in Automation—provides a structured yet flexible approach, adhering to a set of well-defined principles to ensure the successful integration of RE, DevOps, and RPA technologies throughout the automation lifecycle.

By combining Requirements Engineering with DevOps practices, the purpose of FRIDA is to optimize the creation and maintenance of automation systems, fostering solutions that are robust and aligned with business needs. This integration seeks not only to execute automated processes efficiently but also to ensure that automation remains continuous, adaptable, and capable of evolving in response to organizational changes. The model encompasses the entire RPA lifecycle, from initial analysis to ongoing post-deployment support, promoting an integrated and sustainable approach to automation.

The following sections detail the phases of the RPA lifecycle, incorporating the relevant Requirements Engineering criteria and DevOps practices that guide each stage.

For the proposed framework, key phases of the RPA lifecycle are identified. These phases are critical for ensuring project success and are outlined as follows [34]:

• Analysis: At this stage, the focus is on understanding the process that will be automated. The RPA team analyzes the current workflow, identifies repetitive, manual, and time-consuming tasks, and assesses the feasibility of automation. It is important to determine whether the process is stable, well documented, and suitable for RPA.

  ○

  Objective: identify processes that are candidates for automation and ensure that automation will bring value to the business.

  ○

  Results: list of priority processes for automation and a clear vision of the expected benefits.

• Requirement gathering: Here, the team details the requirements of the process that will be automated. This includes understanding the business rules, the systems involved, the exceptions that can occur, and the expected results. Collaboration with end-users and stakeholders is crucial to ensure all aspects of the process are considered.

  ○

  Objective: clearly define what the robot should do and how it should behave in different scenarios.

  ○

  Results: document functional and non-functional requirements, which will serve as a basis for development.

• Design–project development: In this phase, the RPA team creates the automation design, including the robot workflow, decision logic, and integration with existing systems. RPA tools are used to develop the robot, following best programming and automation practices.

  ○

  Objective: build a functional robot that meets the defined requirements.

  ○

  Results: a prototype or initial version of the robot, ready for testing.

• Testing phase: The developed robot is tested in a controlled environment to ensure that it works as expected. This includes unit tests (for each part of the robot), integration tests (to verify interaction with other systems), and acceptance tests (to validate with end-users).

  ○

  Objective: identify and correct errors before deployment into production.

  ○

  Results: robot validated and approved for implementation.

• Deployment and hypercare: After testing, the robot is deployed in the production environment. The hypercare phase is a period of intensive monitoring right after deployment, where the RPA team keeps an eye out for any issues or necessary adjustments.

  ○

  Objective: ensure that the robot functions correctly in production and quickly resolve any problems that arise.

  ○

  Results: robot in stable operation and continuous support during the hypercare period.

• Go-live and Support: At this stage, the robot is fully operational and begins performing its tasks in the production environment. Ongoing support is provided to deal with any issues, updates or process changes. Additionally, the robot’s performance is monitored to ensure it continues to deliver the expected results.

  ○

  Objective: keep automation running efficiently and aligned with business needs.

  ○

  Results: fully functioning automation, with ongoing support and continuous improvement if needed.

Each of these phases has distinct characteristics, as described below:

• Analysis: The primary objective of the analysis phase is to identify opportunities for automation and conduct a thorough examination of the processes to be automated,

  Table 5. RE criteria: analysis.

  RE Criteria: Analysis
  The identification of the processes to be automated and their boundaries.
  The assessment of process complexity and its business impact.
  The evaluation of the technical and economic feasibility of automation.
  The identification of integrations with existing systems.
  The establishment of ROI indicators to justify the project.

  and

  Table 6. DevOps criteria: analysis.

  DevOps Criteria: Analysis
  Collaboration between developers and operations, ensuring clear communication from project inception.
  Capacity planning to evaluate the computational resources required for automation.
  Continuous optimization, identifying areas for improvement from the outset to enhance the efficiency of automated processes.

  .

2.

Requirement gathering is a critical phase where the robot’s actions are detailed,

Table 7. RE criteria: requirement gathering.

RE Criteria: Requirement Gathering
Ensuring the process is structured, repetitive, and governed by clear rules.
Detailing interactions with systems, interfaces, and files.
The specification of performance, security, and compliance requirements.
Alignment with stakeholders to collect specific project requirements.

and

Table 8. DevOps criteria: requirement gathering.

DevOps Criteria: Requirement Gathering
The versioning of scripts and assets to manage changes effectively.
Controlled change management processes for evolving requirements.
Early planning of unit and integration tests.

.

3.

Design–Project development: during this phase, robust and scalable solutions are created,

Table 9. RE criteria: design–project development.

RE Criteria: Design–Project Development
Ensuring resilience to errors or system failures.
Planning solutions for exceptions and legacy system integration issues.
Considering compatibility and scalability requirements for future processes.

and

Table 10. DevOps criteria: design–project development.

DevOps Criteria: Design–Project Development
Modular architecture to facilitate maintenance and scalability.
Implementation of CI/CD pipelines for automation of building, testing, and deploying bots.
Comprehensive documentation and continuous feedback during development.

.

4.

Testing phase: the testing phase aims to validate the robot’s requirements and performance before deployment,

Table 11. RE criteria: testing phase.

RE Criteria: Testing Phase
Planning test scenarios to validate automations.
Defining performance metrics and monitoring methods.

and

Table 12. DevOps criteria: testing phase.

DevOps Criteria: Testing Phase
Implementation of automated testing and production environment simulations.
Ensuring test coverage for use cases, exceptions, and edge conditions.
Planning redundancy solutions and load balancing mechanisms.

.

5.

Deployment and hypercare: this phase focuses on deploying the robot in production and providing ongoing support,

Table 13. RE criteria: deployment and hypercare.

RE Criteria: Deployment and Hypercare
Defining strategies for the robot’s continuous maintenance and updates.
Specifying performance metrics and error reduction strategies.

and

Table 14. DevOps criteria: deployment and hypercare.

DevOps Criteria: Deployment and Hypercare
Implementing real-time monitoring and proactive alerts.
Measuring KPIs such as success rates and average execution times.
Securely managing credentials and auditing bot actions.

.

6.

Go-live and support: The go-live and support phase emphasizes project continuity and evolution,

Table 15. RE criteria: go-live and support.

RE Criteria: Go-Live and Support
Addressing limitations of available RPA tools.
Defining scalability criteria for replicable processes.

and

Table 16. DevOps criteria: go-live and support.

DevOps Criteria: Go-Live and Support
Ensuring consistency across production, development, and testing environments.
Implementing redundancy to maintain continuity during failures.
Monitoring and optimizing bot performance based on established KPIs and metrics.

.

By effectively integrating Requirements Engineering criteria and DevOps practices throughout all phases of the RPA lifecycle, FRIDA seeks to ensure the success and sustainability of automation projects. This structured approach promotes the development of solutions that are robust, efficient, and aligned with business objectives.

4.2. Characteristics and Benefits of the Framework (FRIDA)

FRIDA stands out as a robust and innovative model designed to enhance the development and maintenance of Robotic Process Automation (RPA) systems through the integration of Requirements Engineering (RE) and DevOps practices. Its unique features and associated benefits are outlined below, demonstrating how it delivers value across the automation lifecycle.

Characteristics of the Framework

• Holistic lifecycle approach: FRIDA covers the entire RPA lifecycle, from initial analysis to ongoing support. By embedding RE and DevOps practices in every phase, it ensures that automation projects are planned, implemented, and maintained comprehensively.
• Structured yet flexible design: The framework provides a well-defined structure, while allowing flexibility to adapt to varying project requirements and organizational needs. This balance ensures relevance across industries and project scales.
• Integration of key practices:
  • Requirements engineering: focuses on capturing detailed, business-aligned requirements to guide automation projects effectively.
  • DevOps methodologies: ensure continuous integration, testing, and delivery while fostering collaboration between development and operational teams.
• Emphasis on human–computer interaction (HCI): FRIDA prioritizes creating intuitive and user-friendly systems, optimizing interactions between users and automated processes.
• Modularity and scalability: The framework advocates for modular design principles, ensuring that automation solutions are scalable and maintainable. This approach supports future process enhancements and adaptability to evolving business needs.
• Performance monitoring and continuous feedback: real-time monitoring tools and continuous feedback loops are embedded in the framework to ensure optimal performance and alignment with business objectives.
• Sustainability and evolution: FRIDA is designed to evolve with organizational changes, making automation solutions sustainable over time. This characteristic supports long-term value creation and adaptability.

Benefits of the Framework

• Increased operational efficiency: By streamlining the automation lifecycle, FRIDA reduces development time and operational overhead, enabling organizations to achieve faster time-to-value.
• Improved collaboration: The integration of RE and DevOps fosters enhanced collaboration between stakeholders, including developers, operations teams, and business units. This alignment reduces silos and improves project outcomes.
• Enhanced user experience: the focus on HCI ensures that automation solutions are intuitive and user-friendly, reducing resistance to adoption and improving overall satisfaction.
• Minimized risks and errors: the structured approach of FRIDA, combined with rigorous testing and monitoring practices, mitigates risks associated with automation deployment and maintenance.
• Scalability and futureproofing: the modular design of FRIDA enables organizations to scale automation solutions seamlessly, supporting business growth and future technological advancements.
• Sustainable automation: by promoting adaptability and continuous improvement, the framework ensures that automation solutions remain relevant and valuable, even as organizational needs evolve.
• Data-driven decision-making: through performance monitoring and real-time analytics, FRIDA empowers organizations to make informed decisions about optimizing and expanding their automation initiatives.
• Alignment with business goals: By embedding RE principles, FRIDA ensures that automation projects are closely aligned with strategic objectives, delivering solutions that drive measurable business impact.

FRIDA exemplifies a forward-thinking approach to RPA development, combining the strengths of Requirements Engineering and DevOps to create a comprehensive, adaptable, and effective framework. Its characteristics and benefits make it a valuable tool for organizations aiming to achieve sustainable automation, operational excellence, and enhanced user experiences.

5. Case Study

5.1. Case Study Presentation

The case study focuses on a mid-sized financial services company that specializes in processing loan applications for small and medium-sized enterprises (SMEs). The organization manages an average of 1500 loan applications per month, each requiring meticulous review, cross-referencing with client documentation, and the validation of compliance with financial regulations. The primary challenge lies in the time-consuming and error-prone nature of these processes, which are critical for maintaining regulatory compliance and customer satisfaction.

The organization’s operational team consists of 12 employees, working in three shifts to ensure 24/7 coverage. Despite their efforts, the manual nature of the process has resulted in several challenges:

• Inefficiencies in processing time:
  • Each application takes an average of 2 h to process fully, leading to delays in meeting customer expectations.
  • Bottlenecks often occur during peak periods, such as the end of fiscal quarters.
• High error rates:
  • Manual data entry errors lead to incorrect assessments or rejections of valid applications.
  • Approximately 8% of applications require rework due to inconsistencies in documentation validation.
• Strained employee workloads:
  • Employees report high levels of stress due to repetitive tasks and the pressure to meet tight deadlines.
  • Turnover rates have increased, with 3 employees leaving in the past year.
• Lack of scalability:
  • As the company expands, the current processes struggle to keep up with the growing volume of applications.
  • Management has expressed concerns about the long-term sustainability of the current approach.

To address these issues, the company identified the need for a Robotic Process Automation (RPA) system with integrated human–computer interaction (HCI). This solution would provide the following benefits:

• Automate repetitive tasks such as data entry, document validation, and compliance checks.
• Provide employees with intuitive dashboards to monitor and intervene in processes as needed.
• Reduce processing times while maintaining high accuracy levels.
• Enhance employee satisfaction by allowing them to focus on higher-value tasks.

The case study aims to achieve the following:

• Analyze the inefficiencies and challenges in the current manual processes.
• Evaluate the impact of implementing FRIDA on the development and deployment of an RPA system.
• Compare the performance metrics before and after the framework’s implementation, focusing on the following aspects:
  • Processing time per application.
  • Error rates and rework percentages.
  • Scalability and adaptability of the automation solution.

5.2. Application of the Framework to the Case Study

The implementation of FRIDA in the case study of the financial services company, specialized in processing credit applications for small and medium-sized enterprises (SMEs), aimed to optimize the automation lifecycle, from initial analysis to the continuous maintenance of the automated system. The detailed application of FRIDA across the phases of the RPA lifecycle will be described, with an emphasis on the improvements achieved in terms of efficiency, error reduction, and scalability, based on the established requirements and practices.

• Phase 1: Analysis

The analysis phase, which aimed to identify automation opportunities and assess the feasibility of processes to be automated, began with a thorough assessment of the manual operations carried out by the company’s team. The analysis process was guided by FRIDA criteria, focusing on identifying repetitive, time-consuming, and error-prone tasks, such as manual data entry and document validation. To this end, a detailed review of each step in the credit application processing procedure was carried out, from receipt to final validation.

The analysis revealed that data entry was one of the main sources of inefficiency, with each application taking an average of two hours to process. Additionally, the high error rate of approximately 8%, stemming from inconsistencies in manually entered data, was identified as a critical factor affecting both regulatory compliance and customer satisfaction. Based on this evaluation, the goal was to automate the tasks of data entry and document validation, reducing processing time and human errors.

Furthermore, the assessment of process complexity and the technical and economic feasibility of automation also helped identify the necessary integrations with the company’s legacy systems, such as customer management software and financial compliance verification systems. The creation of return on investment (ROI) indicators was essential to justify the implementation of automation, demonstrating that error reduction and increased efficiency would bring clear financial benefits in both the short and long term.

• Phase 2: Requirement Gathering

In the requirement gathering phase, close collaboration with the company’s stakeholders was undertaken to define the detailed specifications of the processes to be automated. The use of FRIDA ensured that all requirements were aligned with business needs, considering both operational and regulatory demands.

It was specified that the data entry process should be automated to ensure the accuracy and consistency of information, with particular attention to interactions with external systems, such as customer databases and document validation systems. Automation should also ensure that security requirements and compliance with financial standards are met, while also ensuring the scalability necessary to accommodate the company’s growth.

In addition, the requirement gathering phase involved defining performance criteria, such as the maximum processing time per application, as well as the need for an intuitive user interface for monitoring and manual intervention when necessary. Version control practices for scripts and assets, aligned with the framework, ensured that changes were managed and that automation could be easily updated as new requirements or improvements arose.

• Phase 3: Design–Project Development

The design phase was the stage where the automation solution was effectively conceived and developed. Using the principles of the FRIDA, the RPA team created a modular and scalable architecture for the automation robot, with the aim of ensuring that the solution was robust and easily adaptable to future changes. The architecture was designed seamless integration with the company’s existing systems, such as the CRM and compliance verification systems, without disrupting the daily workflow.

One of the key considerations during this phase was the integration of RPA tools with monitoring and feedback systems to ensure that automation results could be continuously evaluated and adjusted, guaranteeing maximum efficiency and regulatory compliance. The modularity of the system also allowed for the easy incorporation of new processes into the automation system in the future, without the need for major reengineering.

• Phase 4: Testing Phase

The testing phase was conducted according to the guidelines set by the FRIDA, ensuring that all performance and compliance requirements were met before the system was launched into production. The RPA team developed and executed a series of test scenarios to validate the functionalities of data entry, document validation, and financial compliance verification. Through rigorous testing, any process failures were identified, and we ensured that the robot operated according to specifications, both in terms of performance and accuracy.

Additionally, the testing phase included simulations of production environments, allowing for an evaluation of the system’s resilience and its ability to handle load peaks, such as those experienced at the end of each fiscal quarter. The test results indicated a significant reduction in the processing time for credit applications, with automation managing to process an application in an average of 20 min, a substantial improvement from the previous two hours.

• Phase 5: Implementation and Hypercare

Following the successful completion of testing, the robot was deployed into production, initiating the hypercare phase, during which its performance was closely monitored and corrective interventions were made if necessary. During this phase, the RPA team closely monitored the system’s behavior, collecting user feedback to adjust the automation and resolve any unexpected issues that arose. Real-time monitoring, aligned with FRIDA practices, enabled a proactive approach to resolve issues before they affected the business continuity.

Moreover, during the hypercare phase, continuous update and maintenance strategies were implemented for the automated system, ensuring that automation remained aligned with regulatory compliance requirements and ever-evolving business needs.

• Phase 6: Go-live and Support

In the go-live phase, the automated system was fully integrated into the company’s daily processes, enabling the complete automation of credit application processing, with minimal human intervention. Automation provided greater consistency in processes and reduced operational errors, allowing the company’s team to focus on higher-value tasks, such as analyzing more complex cases.

The ongoing support phase involved a robust monitoring system, allowing the RPA team to adjust and improve the system as necessary. The key performance indicators (KPIs) defined, such as processing time per application and error rate, were closely monitored to ensure that efficiency and accuracy objectives continued to be met over time.

6. Analysis of Results and Discussion

6.1. Analysis of Results

In this section, the results obtained from the implementation of FRIDA in automating the credit application process are presented and analyzed. The data used for analysis were gathered from the company’s operational records, both before and after automation’s implementation. Data sources include the company’s internal systems, such as the customer management software and financial compliance verification systems, as well as feedback from the RPA team and the stakeholders involved in the project.

The key variables used for analysis were as follows:

• Processing time per application (time spent processing each credit application);
• Error rate (percentage of errors in the application processing);
• Manual vs. automated steps (comparison of tasks performed manually versus automated);
• Compliance rate (the rate at which processed applications met regulatory requirements);
• System scalability (ability to handle increased volumes of applications).

Below is a summary of the collected data, both before and after the implementation of the framework. The values are presented in a balanced manner, reflecting realistic, moderate improvements in the key performance indicators,

Table 17. Key performance indicators.

	Before Implementation (Manual Process)	After Implementation (Automated Process)	Improvement (%)
Processing time per application	120 min	20 min	83%
Error rate	8%	1.5%	81.25%
Manual steps	100% of tasks manually performed	20% of tasks automated	80% reduction
Compliance rate	92%	98%	6.5% improvement
System scalability	Limited scalability (manual capacity)	Highly scalable, handling higher volumes	Significant increase

.

Data collection and sources:

• Processing time per application: Data were collected from the company’s internal CRM system and process logs before and after automation. Time was measured from the receipt of the application to final approval or rejection.
• Error rate: The error rate was tracked through manual records in the credit application system, identifying inconsistencies in data entry and validation before automation. After implementation, the error rate was monitored through the automated system’s logs and feedback from the compliance team.
• Manual vs. automated steps: The analysis of manual and automated processes was conducted based on detailed documentation of work processes and logs generated by RPA (Robotic Process Automation) software. To identify which steps were automated and which remained manual, we used specific tools such as UiPath for automation. This tool was chosen due to its wide adoption in the market and its ability for integration with existing systems, such as the CRM used by the organization. In the case of this study, the CRM in question was Salesforce, a widely used platform for customer relationship management. Salesforce was chosen for its flexibility and customizability, although its native integration with RPA required adaptations due to the lack of specific APIs for automating financial processes, such as loan application processing. To overcome these limitations, DevOps practices were incorporated into the automation workflow, aiming for more efficient integration between RPA and CRM. Tools like Jenkins were used to orchestrate continuous integration (CI/CD) pipelines, enabling test automation and the deployment of RPA scripts in an agile and reliable manner. Additionally, Git was used to establish versions of automation scripts, ensuring traceability and collaboration between development and operations teams. This approach allowed for the creation of an environment where updates to Salesforce or RPA processes could be implemented quickly, with a reduced risk of errors and greater consistency. The integration between RPA and DevOps was not limited to the automation of repetitive tasks, but also included the application of principles such as continuous monitoring and feedback loops. For example, logs generated by UiPath have been consolidated into dashboards in Jenkins, enabling the proactive identification of failures and continuous process optimization. This synergy between RPA and DevOps has not only accelerated loan application processing but also created a foundation for future evolution, where automation can be expanded to other workflows, such as credit analysis and financial reporting. This approach allowed for a clear analysis of the gaps between manual and automated processes, highlighting future opportunities for greater integration between systems and the adoption of more advanced automation practices. The incorporation of DevOps in its strictest sense, with the use of tools such as Jenkins and Git, demonstrates how the organization can evolve towards a more integrated software development model, where the automation of business processes and the continuous delivery of technological solutions go hand in hand.
• Compliance rate: Compliance data were obtained from the financial compliance verification system, which tracks the regulatory requirements for each application. The rate of compliance was calculated based on successful completions of the automated and manual processes.
• System scalability: This was measured based on the volume of applications processed per day. Data were obtained from both manual and automated systems, showing the system’s capacity to handle increased loads after automation.

The results presented here demonstrate significant improvements in key areas such as processing time, error reduction, and scalability. These outcomes are reflective of a balanced and realistic performance boost, aligning with the moderate expectations of the framework implemented.

The bar chart, Figure 2, illustrates the significant improvements in key variables following the implementation of the FRIDA. It compares the performance of the manual process before implementation with the automated process afterward. The data reveal an impressive 81.25% reduction in the error rate, dropping from 8% in the manual process to 1.50% after automation. Furthermore, the automation resulted in a drastic 80% decrease in manual steps, reducing the reliance on human intervention from 100% to 20%. Although the improvement in the compliance rate was more modest, it still showed a notable 6.5% increase, rising from 92% to 98%. These results highlight the substantial efficiency gains and enhanced accuracy achieved through the automation process.

6.2. Discussion

The implementation of FRIDA in automating the credit application process has yielded substantial improvements across several key performance indicators, demonstrating significant gains in efficiency, accuracy, and scalability. However, these outcomes must be contextualized within the organizational environment to fully appreciate their broader implications.

The automation process reduced the average processing time per application by 83%, from 120 min in the manual process to just 20 min post-automation. This marked improvement underscores the efficiency gains facilitated by Robotic Process Automation (RPA). By automating routine tasks, the system can process applications more rapidly, significantly enhancing operational capacity and customer satisfaction. Nevertheless, more complex cases may still require additional processing time, suggesting opportunities for further refinement of the automation framework.

The error rate was reduced by 81.25%, falling from 8% to 1.5%, reflecting the accuracy improvements achieved through automation. By minimizing human errors in data entry, validation, and compliance checks, we reduced the system’s reduced, thereby decreasing rework, enhancing data integrity, and improving overall process quality. Continuous monitoring, however, remains essential to mitigate the risk of new errors emerging as the system evolves.

An 80% reduction in manual steps highlights the framework’s capacity to automate the majority of tasks, enabling employees to focus on more strategic activities. However, the remaining 20% of manual tasks, particularly those involving complex decision-making or exceptions, indicate areas where further automation could enhance efficiency. Exploring machine learning and AI-driven processes may offer solutions to address these residual tasks.

The compliance rate increased by 6.5%, rising from 92% to 98%, demonstrating the system’s ability to consistently apply regulatory requirements and reduce non-compliance risks. Automation ensures that predefined rules are uniformly enforced, improving adherence to regulatory standards. Future work could focus on enhancing the system’s ability to adapt to evolving regulations and introducing more sophisticated compliance checks.

Scalability improved significantly following the framework’s implementation, enabling the system to handle higher application volumes without the constraints of manual processes. This scalability is particularly critical for organizations experiencing growth or fluctuating demand. However, challenges may arise during periods of high application volume or during integration with other systems. Future efforts should aim to enhance the system’s adaptability to such variations.

The integration of Requirements Engineering (RE) and DevOps practices in the development of RPA systems has demonstrated clear benefits in terms of operational efficiency and user experience. Combining these practices provides a structured approach to identifying requirements early in the development lifecycle (through RE) while enabling continuous updates and rapid deployment cycles (through DevOps). This dual approach ensures that RPA systems align closely with user needs, resulting in functional and user-friendly solutions.

However, deploying DevOps in process automation environments presents specific challenges. These include the need for synchronization between development and operations teams, particularly in organizations with traditional cultures, and the need to manage dependencies between legacy systems and new automation solutions. The scalability of CI/CD pipelines in high-volume application scenarios may also require infrastructure adjustments and resource reallocation. These limitations underscore the importance of a flexible and adaptive approach, allowing for continuous refinement of the framework to address real-world complexities.

The application of human–computer interaction (HCI) principles has proven critical in ensuring the intuitive operation of RPA systems. Systems developed with a strong emphasis on HCI design have led to higher user satisfaction and adoption rates. Users interacting with bots that are easy to navigate and understand tend to trust the system more and utilize it more effectively, contributing to the overall success of automation initiatives.

Specific HCI design principles applied in FRIDA include consistency, ensuring that bot interfaces follow predictable and familiar patterns; immediate feedback, providing users with clear and timely responses to their actions; and minimized cognitive load, simplifying interactions to reduce perceived complexity.

Iterative usability testing during development allowed for the early identification and resolution of interaction issues, reducing rework costs and increasing system adoption. This iterative approach, supported by quantitative and qualitative metrics, illustrates how HCI was practically and measurably integrated into the FRIDA, moving beyond superficial discussions.

The findings from the case study further reinforce the hypothesis that integrating RE and DevOps enhances operational efficiency. By reducing development cycle times and enabling continuous delivery, organizations can achieve the faster automation of business processes, resulting in significant time and cost savings.

FRIDA represents an innovative approach to process automation by synergistically integrating three critical disciplines: RE, DevOps, and HCI. While many automation solutions focus on only one of these areas, FRIDA proposes a structured and iterative combination that addresses both technical efficiency and user experience. The innovation lies in how the framework connects these disciplines, creating a continuous development cycle that prioritizes precise requirement identification, the agile delivery of functionalities, and usability assurance.

For instance, integrating RE at the project’s outset ensures that automation requirements are accurately mapped and aligned with user needs. This is complemented by DevOps practices, which the enable rapid implementation of updates and improvements without disrupting existing operations. Additionally, applying HCI principles throughout development ensures that bots are intuitive and effective, increasing user adoption and satisfaction.

This integrated approach is particularly innovative as it addresses a common challenge in automation projects: the disconnect between technical efficiency and user experience. By combining RE, DevOps, and HCI, FRIDA not only enhances operational efficiency but also ensures that automation systems are sustainable, scalable, and aligned with user expectations. This combination of practices, supported by usability metrics and continuous feedback cycles, represents a significant advancement over traditional automation approaches.

However, while the integration of RE, DevOps, and HCI in RPA projects shows considerable promise, certain challenges remain. Despite the improvements observed in the case study, transitioning from manual to automated processes requires careful planning and clear cross-departmental communication. Additionally, ensuring the sustainability and scalability of RPA systems in complex, dynamic environments necessitate continuous monitoring and the iterative refinement of both the RPA bots and the supporting framework.

In conclusion, the implementation of FRIDA has led to substantial improvements in processing time, error reduction, manual steps, compliance, and scalability. These results underscore the transformative potential of RPA in enhancing business processes. Continuous monitoring, refinement, and the expansion of the framework are essential to sustain and enhance these benefits, particularly by addressing remaining manual tasks and potential future challenges. FRIDA represents a significant contribution to the field of process automation, not only due to the practical results achieved but also because of its integrated and innovative approach. By combining RE, DevOps, and HCI in a structured manner, FRIDA addresses critical challenges in terms of efficiency, usability, and scalability, offering a solution that surpasses traditional approaches. The inclusion of usability metrics, modern DevOps tools, and a transparent discussion of limitations and challenges reinforces the rigor and relevance of the study. These elements highlight how FRIDA not only meets current automation needs but also sets a benchmark for future initiatives in the field.

7. Conclusions

This study highlights the significant impact of integrating Requirements Engineering (RE) and DevOps practices into the development of Robotic Process Automation (RPA) systems. By adopting a structured approach to requirement elicitation and management, alongside ensuring continuous integration and delivery, organizations can streamline the RPA development process while enhancing both operational efficiency and user experience.

The emphasis on human–computer interaction (HCI) in RPA system development is fundamental to ensuring that these systems are not only efficient in automating processes but also user-friendly and intuitive, thereby improving overall satisfaction and user adoption. The combination of RE, DevOps, and HCI establishes a robust framework for delivering high-quality, sustainable RPA solutions that align closely with end-user needs and expectations.

The study fully addressed the central research question and both proposed hypotheses. The research demonstrated how the integration of RE and DevOps can optimize the development of RPA systems, directly answering the central question by evidencing improvements in operational efficiency and user experience. Hypothesis H1 was confirmed by the reduction in development cycles and greater predictability and reliability in the delivery of automated solutions, ensuring the continuous and iterative implementation of functionalities.

Furthermore, Hypothesis H2 was addressed by demonstrating that the combination of RE and DevOps improves the user experience by creating more intuitive and easy-to-use RPA systems. The user experience was evaluated from the perspective of interaction with the RPA system, considering factors such as usability, interface fluidity and adaptation of the system to operational needs. Although the same CRM was used before and after the implementation of FRIDA, the introduction of the RE + DevOps framework promoted significant improvements in human–computer interaction. Reducing inconsistencies in automation flows, refining requirements capture, and continuous integration resulted in a more fluid and productive experience for users. Furthermore, the qualitative analysis based on feedback from professionals who interact with the system reinforces that the adoption of this methodology improved users’ perception of the ease of use and effectiveness of automated solutions. Therefore, both the central research question and H1 and H2 were consistently addressed within the case study, with evidence supporting improved operational efficiency and user experience in the context of RPA systems.

Despite the promising findings of this study, several limitations must be considered. Firstly, the case study focused on a specific organization, and the results may not be fully generalizable across all industries or contexts. The success of integrating RE, DevOps, and HCI into RPA projects may vary depending on organizational size, the complexity of automated processes, and the expertise of the development and operations teams involved.

Furthermore, the study primarily examined the initial implementation phase of the framework, with limited exploration of its application over time. Future research could investigate the long-term effectiveness and adaptability of the proposed framework in dynamic environments where business requirements and technological capabilities are continuously evolving.

To strengthen the analysis of the proposed framework’s efficacy, future studies should incorporate user-centered evaluation techniques, such as usability testing and cognitive workload analysis. These methodologies would enable the objective assessment of the impact of integrating RE and DevOps on user experience in RPA systems. Usability testing could provide valuable insights into the intuitiveness and ease of use of automated solutions, while cognitive workload analysis could assess the mental effort required for system interaction. The adoption of these metrics would not only validate improvements in user experience but also guide future refinements in RPA implementation based on the proposed methodology.

Future research should expand upon the findings of this study by exploring the application of the proposed framework across various industries and organizational contexts. Additionally, further investigation into the integration of machine learning and artificial intelligence with RE and DevOps in RPA systems could provide an opportunity to examine how these technologies enhance the intelligence and adaptability of automated processes. By incorporating advanced analytics and predictive capabilities into RPA, organizations can achieve even greater operational efficiencies and improve decision-making processes.

Finally, the development of practical tools and guidelines to support the implementation of RE, DevOps, and HCI practices in RPA projects would be highly beneficial for practitioners seeking to apply these approaches within their organizations.