Swazona

Swazona needed a platform that could handle complex application workflows, document verification via AI, and a massive data model spanning 150 entities with intricate relationships.

Architected the entire database schema with 150 entities using Drizzle ORM. Integrated Amazon Rekognition for face-based attendance and document verification. Built the frontend with React and the backend with Node.js.

Successfully deployed a platform handling thousands of applications. AI verification reduced manual review time by 70%. The 150-entity schema performed well under load with proper indexing.

Large schemas need careful migration planning. Event sourcing would have been valuable for audit trails. Amazon Rekognition is reliable but needs fallback flows for edge cases.