MoneySuperMarket, one of the UK’s leading price comparison platforms, partnered with Firemind and AWS to take a generative AI concept from proof of concept to live production in just six months. The result was Money Concierge, an AI-powered assistant designed to simplify complex credit card decisions, improve customer confidence, and drive stronger engagement at scale.
Challenge
Choosing the right credit card is one of the most complex decisions customers face on comparison platforms.
Customers must navigate:
- Eligibility criteria that vary by provider
- Introductory offers and interest-free periods
- Fees, repayment structures, and long-term costs
- Uncertainty around approval likelihood
For MoneySuperMarket, this complexity created friction in the customer journey and limited conversion opportunities. The challenge was to simplify decision-making without compromising accuracy, trust, or regulatory standards — while ensuring any AI solution could be deployed securely and responsibly at scale.
Solution
MoneySuperMarket worked with Firemind, an AWS Premier Tier Services Partner, to design and deliver Money Concierge, a customer-facing GenAI assistant built entirely on managed and serverless AWS services.
Money Concierge is embedded directly into the credit card comparison journey and provides:
Personalised recommendations based on customer eligibility and financial context
- Clear explanations of card benefits, limitations, and costs
- Custom repayment plans to help customers avoid interest charges
- Natural language interactions that simplify complex financial information
The solution leverages Amazon Bedrock and Retrieval-Augmented Generation (RAG) to ensure responses are accurate, contextual, and grounded in MoneySuperMarket’s own product data and terms.
Responsible AI and data privacy were designed in from day one. Customer data remains within the AWS environment, is encrypted in transit, and is not used to train foundation models.
The Implementation
To accelerate delivery, MoneySuperMarket created a dedicated internal AI team with the autonomy to experiment, iterate, and scale. Firemind supported the programme end to end — from architecture design to production deployment.
Key implementation highlights included:
- Serverless-first architecture using AWS Lambda and Step Functions
- Secure, private model access via Amazon Bedrock and AWS PrivateLink
- Knowledge Bases with RAG, backed by Amazon OpenSearch Serverless
- Highly available, modular design spanning multiple Availability Zones
This approach allowed the team to remove infrastructure overhead, focus on customer experience, and move from PoC to production in just six months.
The Results
- Quantitative Outcomes
- 168,000 customers supported in their credit card journey
- 96% positive customer feedback
- Faster access to relevant, eligibility-based credit card options
- Customer Experience Impact
- Clearer understanding of credit card products and trade-offs
- Increased confidence in financial decision-making
- Reduced friction in a traditionally complex journey
- Operational & Strategic Impact
- Proven path from GenAI experimentation to production
- Scalable foundation for extending AI across other financial products
- Strong alignment with responsible AI and data governance standards
Conclusion & Key Takeaways
This project demonstrates how generative AI can deliver real customer and business value when paired with the right strategy, architecture, and delivery partner.
Start with the customer problem, not the technology
Dedicated AI teams accelerate innovation and reduce time to value
Managed and serverless services enable speed, security, and scale
Responsible AI by design builds trust in customer-facing solutions
By working with Firemind and AWS, MoneySuperMarket successfully moved beyond experimentation — turning GenAI into a live, trusted product used by hundreds of thousands of customers.
