MoneySuperMarket (MSM), the UK’s leading price comparison platform for financial services, partnered with Firemind to transform its customer experience using AI-driven insurance insights.
By deploying a multi-agent orchestration framework powered by Amazon Bedrock and Claude LLMs, MSM achieved 90% accuracy in data-driven insights and reduced response latency to under 30 seconds.
The solution provides users with transparent, personalised explanations for their insurance quotes, improving trust, reducing drop-offs, and setting a new standard for customer engagement in the comparison market.
Challenge
The price comparison market is defined by low loyalty and high price sensitivity: customers frequently switch between platforms seeking the cheapest quote.
MSM’s strategic objective was to increase user trust and quote conversions by explaining why a premium costs what it does. However, several obstacles stood in the way:
- Fragmented and complex data sources (quotes, MOT records, policies, market stats).
- A lack of transparency in price reasoning leading to user uncertainty.
- Long latency in generating tailored insights.
- The need for a scalable, low-maintenance solution adaptable across insurance types.
To stay competitive, MSM needed a way to turn technical data into human, contextual insight, delivered instantly and accurately.
Solution
Firemind designed and implemented a GenAI-powered, multi-agent system that processes live user and market data to generate tailored insurance insights.
Key Features
- Multi-Agent Orchestrator: Coordinates AI agents responsible for quote analysis, statistics, and contextual reasoning.
- Data Retrieval Layer: Integrates with MSM APIs to unify MOT data, user profiles, and insurance quotes.
- LLM Agents on Amazon Bedrock: Specialised agents (e.g., Quote Comparison, Statistics, MOT Analysis) reason independently before outputs are unified.
- Aggregator Layer: Delivers HTML-ready, human-like summaries that fit directly into MSM’s user interface.
This approach allows MSM to provide personalised, explainable insights in seconds transforming how customers interpret insurance data.
The Implementation
The project was rolled out in two phases:
- Motor Insurance (Phase 1): Established the end-to-end agent pipeline to analyse user quotes, MOT history, and market benchmarks.
- Home Insurance (Phase 2): Adapted workflows to handle property data, risk assessment, and regional averages.
The technical setup included:
- APIs for real-time data retrieval and AI analysis.
- Parallel data ingestion for efficiency and fault tolerance.
- Deployment via Terraform and GitHub Actions, ensuring reproducible environments.
Average total latency: 20–30 seconds, including full data retrieval and LLM reasoning.
Each session invoked up to four agents, each fine-tuned for consistent, low-temperature outputs and minimal hallucination.
The Results
- Quantifiable Outcomes
- 90% accuracy in merging and analysing multi-source data.
- Up to 30% faster insight generation, compared to traditional data workflows.
- 100% compliance with MSM’s content standards through structured output validation.
- 20–30 second total latency, enabling near real-time feedback to customers.
- Customer Experience Impact
- Transparent breakdowns explaining insurance pricing in natural language.
- Clear, data-backed recommendations such as: “Your premium is below the average for your region and age group. Increasing your voluntary excess could save £50.”
- Improved user trust and retention, reducing the need to seek competitor comparisons.
- Operational Benefits
- Reduced manual data analysis and validation time.
- Scalable infrastructure supporting both motor and home insurance — with potential expansion to travel and pet insurance.
- Built-in performance tracking via DynamoDB and CloudWatch for ongoing optimisation.
Conclusion & Key Takeaways
Firemind’s collaboration with MoneySuperMarket demonstrates the power of AI agents and GenAI orchestration in enhancing customer engagement and transparency.
Trust drives conversion
Speed and scalability matter
Modular AI workflows
This project has laid the foundation for MSM’s future in intelligent, explainable price comparison, blending advanced AI with a customer-first philosophy.
