R.E.A.L.M.: Revolutionizing Project Risk Assessment

Inspiration

Working with Indonesia's largest reinsurance company opened my eyes to the complexities of mega-project risk assessment. The challenge of balancing qualitative and quantitative approaches in scoring these projects was particularly intriguing. We implemented a 5C approach (Character, Capacity, Capital, Collateral, and Conditions) to assess projects, but I realized there was potential for a more streamlined, tech-driven solution.

What We Learned

Through developing R.E.A.L.M., we gained deep insights into:

The intricacies of project-based facultative reinsurance
The power of combining traditional risk assessment methods with AI
The importance of user-friendly interfaces in complex financial tools
The challenges of integrating qualitative data into quantitative models

How We Built It

We built R.E.A.L.M. using a stack that includes:

Streamlit for the frontend, ensuring a responsive and intuitive user interface
Python for the backend, handling complex calculations and data processing
Perplexity API exploring the internet
SQLite for storage

The core of R.E.A.L.M. is our innovative scoring system that digitizes the 5C approach:

Character: AI-driven analysis of the project management team's track record
Capacity: Automated assessment of project feasibility and timeline
Capital: Real-time financial modeling and stress testing
Collateral: Smart evaluation of project assets and guarantees
Conditions: AI-powered analysis of market conditions and external factors

Challenges We Faced

Our journey wasn't without obstacles:

User Experience: Balancing comprehensive analysis with user-friendliness was tricky. To refine our interface, we conducted multiple user testing sessions.
Processing Time: A technical hurdle was ensuring quick response times while processing complex risk calculations. To address this, we optimized our backend and implemented efficient caching mechanisms.

Despite these challenges, R.E.A.L.M. emerged as a powerful tool that will transform project-based facultative reinsurance assessment, making it faster, more accurate, and more accessible to underwriters worldwide.