B.Tech CS student in India. I build backend services, AI agents and ML systems, and I contributed to Liquid Galaxy as a Google Summer of Code 2026 contributor.
Looking for backend, ML or systems internships from November 2026.
Google Summer of Code 2026, Liquid Galaxy — worked on LiquidGalaxyLAB/lg-rpg, a real-time multiplayer RPG rendered across the screens of a Liquid Galaxy rig, with players joining from a Flutter controller app on their phones (Node.js, Socket.IO, Phaser, Flutter). 49 merged pull requests, including the objectives and rewards system and the controller pages.
Most of my larger work lives in private repositories. Happy to walk through the code on a call or share access for a hiring process.
| Project | What it is | Stack |
|---|---|---|
| Zenith Agent · private | Autonomous coding agent. Reads and edits a codebase, runs code in Docker sandboxes, gates risky shell commands behind approval, and delegates work to sub-agents. Works with Gemini, OpenRouter and local Ollama models. | Python, Go, Docker |
| InView · private | Face-recognition attendance system built as microservices. RTSP camera streams feed a recognition service; auth, attendance and streaming are separate FastAPI services with their own Postgres databases, joined by RabbitMQ. Deploys to Kubernetes (EKS). | FastAPI, React, PostgreSQL, RabbitMQ, Kubernetes |
| SprintStack · private | Runs my interview prep on GitHub Actions and a Cloudflare Worker with no server. Syncs LeetCode solutions, checks daily activity, sends a weekly digest, and uses Gemini to suggest problems from weak topics through a Telegram bot. | JavaScript, Python, GitHub Actions, Cloudflare Workers |
| Interview Agent | Voice-based technical interviewer. Reads your resume, asks questions out loud, listens, and writes an evaluation report at the end. | Python, AWS Bedrock, AWS Polly, Sarvam STT |
| SeismoSphere | Real-time earthquake map for Liquid Galaxy using the USGS feed. Filter by magnitude and time window, fly the camera to an epicentre. | Flutter, Riverpod, SSH |
| VitaLens | Scan a packaged food label and get a verdict based on your own health profile. | TypeScript, Gemini |
| Face Detection | Face detector on a VGG19 backbone with a classification head and a bounding-box head, trained on an augmented custom dataset, running live on webcam. | TensorFlow, OpenCV |
- Languages: Python, TypeScript/JavaScript, Java, Go, Dart
- Backend: FastAPI, Node.js, PostgreSQL, Redis, RabbitMQ
- ML: PyTorch, TensorFlow, scikit-learn, OpenCV
- Infra: Docker, Kubernetes, AWS, GitHub Actions, Linux