Challenges

Fragmented Verification Ecosystem :
The banking client faced the difficulty of manually aggregating data from over 15+ disparate sources—ranging from government databases (PAN, Aadhaar, GST, CIN) to digital footprint providers (social media, email, phone)—creating a slow and disjointed onboarding process

Event-Driven API Gateway (Kafka & Node.js) :
We utilized Apache Kafka to handle the high throughput of verification requests. When a user submits an application, Node.js pushes an event to Kafka, which triggers parallel workers to fetch data from Aadhaar, PAN, and Social Media APIs simultaneously. This reduced total verification time by processing checks concurrently rather than sequentially.

Dynamic Rules Engine (Redis & Postgres) :
We implemented a custom Rules Engine backed by Redis for speed. It allows ATNA to configure "Risk Score" thresholds instantly. PostgreSQL handles the complex relational mapping between Banking Clients, their Pricing Tiers, and the specific APIs they are allowed to access

Secure Document Handling (MinIO): For standard document uploads (Address proofs, Incorporation certificates), we used MinIO as a secure, S3-compatible object store, ensuring that sensitive file data is isolated from the application logic.