Vyshnavi Pastam

AI Criminal Network Analysis

Smart India Hackathon team project for analyzing criminal-network data and mapping relationships between individuals, entities and activities, using AI-assisted analysis and graph-based relationship mapping. My part: built the FastAPI backend foundation, including the API layer, environment-based configuration, PostgreSQL and Neo4j connection layers, Pydantic validation schemas and a pytest setup. Structured it so route handlers stay separate from database clients, which keeps new workflows easy to add.

Sentinel Gate

Team project. IoT-based access-control system on an ESP32 with RFID authentication, keypad PIN verification, motion detection and a servo-controlled gate. Events are logged to Firebase, with intrusion detection and brute-force attempt monitoring. My part: designed the SAP CPI integration layer to show how the system scales to enterprise use

SAP CPI integration scenarios

Built 7 hands-on SAP Cloud Integration scenarios, including a Groovy script that enriches message payloads. Doing the rest of the series as exam prep.

Rukk

In development: AI-assisted travel planning web app with place search, route directions and trip suggestions. React, Vite and TypeScript front end with an Express and TypeScript backend. Integrates OLA Maps for geocoding and directions, Gemini for AI trip suggestions and Pexels for destination images. All API keys stay server-side behind backend proxy routes. Includes debounced autocomplete with location bias and a straight-line fallback when routing fails.