Case Studies | FinTech

A retrieval-augmented AI webchat for car care

Website AI Chatbot

About

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

Challenge

The challenge had three focus areas.

Operational efficiency at the point of enquiry

Customer enquiries about mechanical problems took time and staff resource to handle. The client needed a way to answer common questions automatically, reducing the effort required while keeping answers helpful and accurate.

Accuracy grounded in existing knowledge

A generic model was not enough. Answers had to be drawn from the client's own documents, so the system needed retrieval-augmented generation: finding the right source material first, then using it to shape the response. Success was measured by comparing the model's answers against the information in the client's documentation.

Validating the platform

Beyond the chatbot itself, the engagement had to demonstrate that Amazon Bedrock and its supporting services were a suitable foundation for retrieval-augmented AI, giving the client confidence to build on it.

Solution

The client's documentation is stored in Amazon S3 as the knowledge base and indexed into a vector database using Amazon OpenSearch Service, so that the most relevant passages for any question can be found quickly by similarity search rather than keyword matching. When a customer asks a question, Amazon Lex interprets the input and identifies intent, and an AWS Lambda function orchestrates the flow between services.

6

AWS services orchestrated into a single retrieval-augmented webchat backend

RAG

Every answer grounded in the client's own documentation, not generic model knowledge

By the numbers:

  • 6 - AWS services orchestrated into a single retrieval-augmented webchat backend
  • RAG - Every answer grounded in the client's own documentation, not generic model knowledge
Changes

The engagement delivered the backend for a knowledge-grounded AI webchat as defined in the project success criteria: successful integration of the webchat service with Amazon Bedrock, information retrieval from the client's documents, and a demonstration that the AWS platform is suitable for taking the service forward. Accuracy was assessed by comparing the model's responses against the client's source documentation, and acceptance followed a successful showcase to the client.

  • Knowledge-grounded answers — A retrieval-augmented pipeline retrieves the most relevant passages from the client's documents before generating each response, so answers stay accurate and on-topic.
  • Conversational context — Amazon DynamoDB maintains conversational memory so that follow-up questions are understood in context.
  • Platform validated — The proof of concept demonstrated that Amazon Bedrock, with OpenSearch, Lex and Lambda, is a suitable foundation for the client's retrieval-augmented AI.
  • Handover — Documentation and a knowledge-transfer session were provided for the client's technical team, with the solution ready to deploy into the client's AWS environment.

With a validated Bedrock-based foundation in place, the client is positioned to move the webchat from proof of concept towards production, extending its knowledge base and integrating it into customer-facing channels.

AWS Stack

Amazon Bedrock

For large language model response generation grounded in retrieved knowledge.

Amazon OpenSearch Service

As the vector database for fast, relevant retrieval.

Amazon S3

As the knowledge base store for the client's documentation.

Amazon DynamoDB

For conversational memory and session context.

Amazon Lex

For interpreting customer queries and intent.

AWS Lambda

For orchestrating retrieval, memory and response generation.

YOU MIGHT LIKE

Related success stories

View all case studies

Case Studies | Insights

Utilising Language Recognition, Speed, and Enhanced Security to Make Social Media a Force for Good

  • Here, we take a detailed look at how the Cloud Combinator team collaborated with another cutting-edge AI service provider that provides intelligent systems to “make social media more social” for brands and users alike.
  • Arwen AI is a UK-based startup specialising in AI solutions to manage and enhance brands’ social media interactions. Founded in 2020 by Matt McGrory, Dr. David Cole, and Joel Bailey, Arwen. AI focuses on using AI to automatically detect and remove spam, toxic comments, and other unwanted content from social media platforms.
  • The team at Arwen have three core products. ‘Moderate’ is focused on identifying and removing toxic content from social media channels. ‘Engage’ helps brands identify and engage with meaningful conversations on social media, and ‘Customize’ allows brands to apply bespoke algorithms to their channels - creating an even more effective moderation and engagement.
Read more
CONTACT US

Ready to turn AI into impact?

We'll help you spot the highest-value opportunities, reduce risk around your first AI initiative, and define a clear path to results from day one.

Why talk to us:

Outcome-driven recommendations

AWS-recognised delivery expertise

Risk-aware AI adoption

Clear next step, not a sales pitch

Start with a focused 20-minute conversation about your goals — no pressure, no commitment.

We use cookies

This website uses cookies to enhance user experience and to analyze performance and traffic on our website.

See our Privacy Policy for details.