Turning community data into instant answers

A SaaS B2B business working with Cloud Combinator on AWS. The client is anonymised at their request.

The challenge had three focus areas, each of which had to be addressed for the application to succeed within the client's secure AWS environment.

At the heart of the solution is a retrieval-augmented generation pattern. The client's reference documents are held in Amazon S3 as the knowledge base data source and indexed into an Amazon OpenSearch Service vector database. When an user asks a question, the relevant passages are retrieved by similarity search and passed, along with the conversation history, to a large language model running on Amazon Bedrock, which generates a grounded, contextually relevant answer.

Cloud Combinator delivered a fully functional, LLM-centric application operating within the client's secure AWS environment, meeting the success criteria set out at the start of the engagement and closely replicating the front-end design the client had specified. The build was signed off against its original scope, with the AWS environment left suitable for the client's team to continue development.

• Grounded, conversational insight — Client questions are answered from the client's own knowledge base through retrieval-augmented generation, keeping responses relevant and rooted in real data.
• Security by design — The solution runs entirely inside the client's AWS account, with IAM-controlled access to storage, Amazon Cognito authentication for users and content delivered securely through Amazon CloudFront.
• Reduced platform dependency — By building the application natively on Amazon Bedrock, the client cut its reliance on external platforms and brought the capability in-house.
• A responsive self-service front end — A React interface, built to the client's mock-ups, lets non-technical users converse with the chatbot while conversational memory maintains context across the exchange.
• Handover — Cloud Combinator provided knowledge transfer and left the client with a tuned, documented AWS environment ready for ongoing enhancement.

With a secure, Bedrock-powered foundation now in place, the client is positioned to extend the application further, deepening the insight it offers clients and building on a platform that is ready for whatever it takes on next.