In a previous article we set out the opportunities for financial institutions to accelerate the speed to market of digital services by leveraging generative AI (GenAI) across the software development lifecycle (SDLC). One such opportunity is accelerating the definition of business requirements – the initial set of activities conducted to identify, elaborate and document changes and enhancements to digital services. 

These activities have traditionally required a high degree of human effort and interaction, facilitated by business analysts (BAs). With the introduction of AI-powered    assistants, the role of BAs can be augmented to improve efficiency and accuracy of requirements definition – ultimately resulting in enhanced speed to market while enabling BAs to focus on strategic, differentiating activities.

In a typical SDLC, BAs act as intermediaries between business stakeholders and technology teams, ensuring that business needs are accurately captured and translated into specifications. They elicit detailed business requirements through various methods, such as interviews, workshops and data analysis. 

BAs then analyze, elaborate upon, and document these requirements, creating comprehensive artifacts such as epics, features, user stories, and test cases. This meticulous process ensures that all stakeholder needs are adequately captured and provides a common understanding around which product development teams can organize.

Defining business requirements can therefore be a laborious task, and the potential for misunderstandings can be high, resulting in frequent back and forth with business stakeholders. Adhering to industry best practices for documenting requirements further adds to the level of effort and cognitive load and can often distract BAs from their primary task of understanding and elaborating business needs.

 At Capco, we believe that requirements definition activities can be greatly accelerated by using GenAI. To validate this value proposition, Capco has developed a GenAI agent – BA Genie – to augment the role of BAs in the SDLC and perform many of the activities associated with the requirements definition process.

Core Features of the BA Genie:

• Context Analysis – BA Genie can consume any format of project documentation to use as context to generate requirements from. It is also able to utilize prerecorded audio/video conversations as input.
  
• Generation of Work Items – BA Genie is configured to break down the provided context and generate multiple high-quality epics, features and user stories using this data. For each of the work item types, it utilizes industry best practices and consider elements specific to financial services such as regulatory compliance and risk mitigation.
• Test Case Generation – Once work items are generated, BA Genie can generate granular test cases to validate the functionality described within, providing test pre-conditions, post-conditions and test steps to be performed.
  
• Integration with Agile Tooling  –  BA Genie seamlessly integrates with popular tools like Azure DevOps and Jira to ensure that generated work items are stored within an organization’s official systems of record.
  
• Refinement of Existing Work Items – BA Genie is also able to augment existing work items and improve them, and generate additional detail based on supplementary documentation and context provided.
  
• Technology Stack Agnostic – BA Genie utilizes a variety of GenAI technologies under the hood but is fully capable of being deployed on an organization’s approved technology stack, or within their private Cloud using native capabilities like AWS Bedrock, Azure AI or Google Cloud Platform Vertex.
  
• Accuracy and Relevance – BA Genie employs a variety of prompting techniques and guardrails to prevent hallucinations and ensure accurate and relevant outputs. It also maintains a human-in-the-loop for validation and enables human actors to edit and refine any outputs produced.

Initial benchmarking of the BA Genie tool confirms the following benefits.

Improved Efficiency. With the automation of a myriad of manual tasks in the requirements definition process, speed to market of new features and changes is greatly enhanced. The requirements process can be accelerated by 30-50% compared to using just BAs, depending on complexity of context provided. BAs also benefit from heightened productivity, with GenAI performing the heavy lifting. 

Enhanced Quality. With common standards baked into BA Genie’s configuration, every work item generated adheres to a robust template and automatically incorporates essential data elements depending on context provided. As a result, the generated work items are more consistent and provide a rich level of detail.

Elevated Roles. With most of the work being done by the BA Genie, the role of the human BA is elevated to that of validation and refinement, and their cognitive capacities are freed up to focus on having better conversations, empathy and engagement with business stakeholder and developing strategic product roadmaps.

CONCLUSION

Requirements definition and analysis has traditionally been human effort intensive and considered a ‘long pole’ in efforts to accelerate speed to market of new capabilities. Capco’s BA Genie provides a practical option for applying GenAI technologies within existing ecosystems and realizing tangible productivity benefits by greatly accelerating business requirements definition activities. Its flexible design makes it possible to implement it within existing ecosystems, lowering the barrier to entry for GenAI adoption.

Contact us for a live demo of the BA Genie, and to discuss how GenAI can accelerate speed to market within your organization.