Selected system · L.E.G.A.L. Intelligence Survey

The system combines a Node.js/Express backend, MongoDB persistence, internal applications, and schema-constrained AI analysis pipelines. I developed the surrounding reliability and security layer for AI inference, including input preparation, response validation, authentication, error handling, and controlled downstream processing.

What moved the work forward

• Architected and shipped a production research and analysis platform combining structured data collection, operational tooling, and AI-assisted qualitative and quantitative analysis.
• Built and maintained a Node.js/Express backend on Azure, integrated with MongoDB, internal applications, and LLM-based analysis services.
• Designed schema-constrained AI pipelines that transform collected information into structured outputs for downstream analytics and human review.
• Developed the AI inference reliability and security layer, including input preparation, response validation, application-level error handling, authentication, and controlled downstream processing.
• Created a separate Tauri/React operations application for internal workflow management, preserving separation from externally accessible systems.