Virtusa Recognized as Global Leader in Gen AI Services by ISG Provider Lens® - Read report
Our client, a large life insurance and mutual life insurance company in the United States, operates a large-scale microservices ecosystem that supports its Platform-as-a-Service (PaaS) infrastructure. As the platform matured, maintaining all the API documentation and ensuring seamless onboarding for new team members became increasingly complex.
Virtusa developed ‘API Assistant,’ an AWS genAI-powered solution leveraging Claude 2.1, Amazon Bedrock, and LangChain to address these challenges and create a centralized, intelligent knowledge base. This conversational AI assistant streamlines information retrieval, enhances platform accessibility, and simplifies onboarding, making API management more efficient and user-friendly.
The challenge
As the client’s microservices ecosystem expanded, maintaining and navigating all its API documentation became increasingly complex.
Key challenges included:
- Lack of a centralized reference for API specifications and platform standards.
- Onboarding inefficiencies, making it difficult for new team members to understand system conventions.
- Limited visibility into request-response structures, slowing down development and troubleshooting.
- Scattered documentation sources, including Swagger files, Excel sheets, PDFs, and Confluence pages, make retrieving relevant information difficult.
The client needed an intelligent solution to provide real-time access to API details, improve documentation visibility, and streamline onboarding.
The solution
Virtusa developed API Assistant, an AWS generative AI-powered application designed to revolutionize user interaction with API documentation.
Built on Claude 2.1 and powered by Amazon Bedrock, the solution utilizes Retrieval-Augmented Generation (RAG) with LangChain to ensure accurate, real-time responses from secure, structured data sources.
Key features of the solution include:
- Automated data ingestion – API documentation is continuously collected from Swagger files, Word documents, JSON files, Confluence, and other sources through a scheduled ingestion process.
- Intelligent information retrieval – Generative AI ensures users receive the most relevant API details based on their queries.
- Conversational AI interface – Users can interact with the assistant naturally, asking detailed questions and receiving instant, context-aware responses.
- Session memory & history tracking – The assistant retains context from previous interactions, allowing seamless, long-form conversations.
- Secure and scalable deployment – Hosted within the client’s Amazon Virtual Private Cloud (VPC) and utilizing Amazon EC2, Amazon S3, and Amazon Bedrock to ensure data integrity, security, and scalability.
By integrating AWS genAI within the platform’s infrastructure, the API Assistant eliminates documentation bottlenecks, accelerates onboarding, and provides an intuitive way for developers to interact with API knowledge.
The benefit
The implementation of API Assistant delivered measurable improvements in API knowledge management and platform efficiency:
- Centralized knowledge base – Consolidated all API details and platform standards into a single, AI-powered reference point.
- Faster information retrieval – Generative AI significantly reduced the time spent searching for API specifications.
- Improved onboarding process – New team members quickly gain knowledge of platform conventions and API structures.
- Enhanced user experience – The conversational interface makes API documentation more accessible and easier to navigate.
- Scalability and security – The solution fully integrates within the client’s secure cloud infrastructure, ensuring compliance and data integrity.
With Virtusa’s AWS genAI-driven API Assistant, the client has transformed API knowledge management, enabling teams to work more efficiently and effectively within their microservices ecosystem.