Software Developer | Backend & System Design Enthusiast | NIT Rourkela
Focused on building scalable web applications and optimizing backend systems. Passionate about modern architecture, clean code, and performance.
- 💻 Core Stack: Java, Spring Boot, React, Next.js, Python
- 🎓 Education: B.Tech in Civil Engineering @ NIT Rourkela ('27)
- 🏎️ Off-screen: Avid F1 fan, AoE2 strategist, amateur photographer, and speed chess enthusiast.
- lichess-org/chess-openings – Contributed 3 merged pull requests improving opening classification and transposition mapping for Lichess, one of the largest open-source chess platforms.
- fitztrev/rosen-score & Project Rosen – Implemented a "Both" platform option to aggregate player statistics from Lichess and Chess.com simultaneously. Maintained a distinct fork featuring a UI overhaul and dynamic PNG generation for downloadable FUT-style stat cards.
- IAmTomShaw/f1-race-replay – Improved project documentation by adding a first-run notice detailing FastF1 cache generation and initial telemetry data loading expectations, streamlining the setup process for new users.
- High-performance, containerized rate-limiting microservice built with Spring Boot to manage API traffic and prevent abuse.
- Implemented the Token Bucket algorithm via Redis to ensure sub-millisecond latency for traffic evaluation across distributed nodes.
- Decoupled the analytics workload by streaming audit logs asynchronously to Apache Kafka, maintaining a strictly non-blocking critical request path.
- Orchestrated a 6-container Docker network featuring PostgreSQL for tier management and a Prometheus/Grafana pipeline for real-time observability.
- Gmail extension (Chrome Manifest V3) and Spring Boot backend that drafts email replies with Google Gemini, adding a Generate Reply button with selectable tones directly inside Gmail's compose window.
- Built a "My Tone" mode using retrieval-augmented generation: embeds a user's past emails (768-dimension Gemini embeddings), retrieves the 5 most similar by cosine similarity, and feeds them to the model as few-shot examples so drafts match the user's own writing style.
- Hardened the Gemini integration with retries and backoff on 429/500/503 errors, an automatic fallback model, and clear upstream error reporting; API keys stay server-side in environment variables.
- Stack: Java 21, Spring Boot, React, Chrome Extension APIs, Gemini API.
- Developed and deployed the official landing platform for an indie game development studio using Next.js 14, React 19, Framer Motion, and Lenis for smooth scrolling.
- Leveraged Next.js Server Components to optimize frontend rendering performance, building dynamic game showcases and interactive news timelines.
- Configured custom domains, managed DNS routing, and continuously optimized UI/UX to support the studio's growing online presence.
- Chess opening analysis system built for structured representation of complex move sequences and variations.
- Engineered an optimized data model recognizing over 3,700 unique openings by integrating the complete Lichess ECO dataset.
- Designed system workflows for automated opening classification, FEN-indexed position generation, and engine evaluation processing.
- Languages: Java, JavaScript, TypeScript, Python, SQL
- Backend & Systems: Spring Boot, Node.js, REST APIs, System Design
- Data & Streaming: PostgreSQL, MongoDB, Redis, Apache Kafka
- AI & Integrations: Gemini API, RAG (embeddings + vector similarity search), Chrome Extensions (Manifest V3)
- Frontend: React, Next.js, Tailwind CSS
- DevOps & Tools: Docker, Git, Prometheus, Grafana, Postman