AWS re:Invent 2025 - Learn how Riskspan is transforming private credit market using GenAI (SMB205)

Transforming Private Credit Market with Generative AI: Riskspan's Journey on AWS

Introduction to Riskspan and Private Credit Market

• Riskspan is a technology company that provides analysis for complex investment asset classes
• Riskspan is 100% cloud-based, running on AWS for over 15 years
• Private credit refers to non-bank entities providing loans to individuals or companies, a $14 trillion market with $9 trillion in untapped assets

Challenges in the Private Credit Market

• Private credit investments are highly customized, with information buried in unstructured data like documents, emails, and handwritten notes
• Analyzing these investments is a manual, time-consuming process, taking 3-4 weeks for analysts to review documents and create Excel models
• Investors need to make decisions quickly, sometimes in less than 10 minutes, but lack access to timely, structured data
• This has led to major risk and losses in the private credit space due to lack of visibility and monitoring

Riskspan's Generative AI Journey on AWS

• Riskspan used the AWS Press Release and FAQ (PRFAQ) process to define their vision and customer-centric solution
• Their architecture leverages AWS services for secure data storage, unstructured data extraction, and large-scale processing using embeddings and Bedrock
• Riskspan used serverless technologies to dynamically generate code and models, achieving 10x scalability and 90x cost reduction compared to manual processes

Lessons Learned in Productizing Generative AI

• LLMs alone are not enough - Riskspan had to provide domain-specific context and break down problems into smaller, structured components
• Structured data and outputs are key for accuracy and reliability of the generative AI models
• Incremental improvements, like creating specialized agents and orchestrating them, were crucial for robustness and continuous learning

Riskspan's Future Vision and Impact

• Riskspan is transitioning to an agent-based architecture using tools like AWS Agent Core to solve specific problems more dynamically
• This allows them to create structures more quickly based on market conditions, provide real-time risk analysis, and enable more continuous learning
• Riskspan has achieved significant business impact, including 8-7% faster customer onboarding, 3-4 week deal evaluation reduced to 3-5 days, and 90% cost reduction

Conclusion

Riskspan's journey demonstrates the power of generative AI in transforming complex, data-intensive industries like private credit. By leveraging AWS services and adopting a structured, incremental approach, Riskspan has been able to unlock a $9 trillion market opportunity and deliver tangible business results for its customers.