Vaibhav
Pawar

Ride booking backend built with six independent microservices communicating exclusively via Kafka events, no direct service-to-service calls. Implements choreography-based Saga pattern where each service reacts to events and handles its own compensating rollback on failure. Driver matching uses Redis geospatial indexing to find the nearest available driver in milliseconds. Real-time GPS tracking streams driver location to the rider via WebSocket and Redis pub/sub. Payments are idempotency-keyed to prevent double charging on duplicate Kafka delivery. Guaranteed event delivery achieved through the Outbox pattern, events survive service crashes. Includes Dead Letter Queue with exponential backoff retry for failed notifications.

REST API connecting employers and jobseekers at scale. Employers create companies, post and manage job listings, and review applicants. Jobseekers search and filter jobs dynamically, apply with resume upload, and track application status in real time. Engineered with a three-layer security model — JWT authentication, role-based access control, and resource-level ownership checks, preventing unauthorised data access at every layer. Features dynamic job search via JPA Specifications, versioned schema management with Flyway, and a fully containerised one-command Docker Compose deployment.

Automated extraction from Random User Generator API, ingested into Kafka streams for real-time processing. Used Apache Spark for transformation and analysis. Processed data stored in Cassandra, orchestrated via Airflow, and containerised with Docker.

Real-time data pipeline for streaming stock data using Apache Kafka, Python, and AWS. Producer streams stock data to a consumer storing JSON records in S3. Integrated AWS Glue for data cataloging and Athena for SQL queries enabling continuous analysis.

AI-powered inventory management system featuring real-time stock monitoring and automated reordering. Certificate of Appreciation at Walmart Hackathon 2024.

Developed and deployed a plant disease detection system using deep learning to identify crop diseases at scale with high accuracy.

Cybersecurity tool for anomaly detection in network traffic and enhanced threat prevention using machine learning algorithms.