I'm a Full-Stack Developer and Computer Science student (2023–2027) based in Noida. My work focuses on building backend infrastructure and autonomous systems from the ground up.
I prefer writing core logic—like custom neural network topologies or probabilistic data structures—over just wrapping high-level libraries
Languages: JavaScript (ES6+), SQL, C++, Python Backend: Node.js, Express.js, REST APIs, WebSockets Database: MongoDB AI/ML: Autonomous AI Agents, MCP servers, TensorFlow.js, Genetic Algorithms Tools: Postman, Git, GitHub, FFmpeg, ImageKit
Built a threat detection pipeline that processes 650K+ URL signatures with ~0.05 ms lookup latency.
- Uses In-memory Bloom Filters (FNV-1a and MurmurHash2) for fast initial checks.
- Integrated an ONNX ML engine directly into Node.js for real-time zero-day classification.
- Optimized with an LRU caching layer to handle high-traffic scanning.
A reinforcement learning simulation built in vanilla JavaScript without PyTorch or TensorFlow.
- Implemented a custom 5-12-4 feedforward neural network with Tanh activation.
- Optimized agent behavior using a 1+ genetic algorithm with Gaussian mutation (
$\sigma=0.2$ ). - Features a procedural traffic engine for testing generalization across stochastic environments.
A mobile-first video feed built with Next.js and TailwindCSS.
- Handles secure auth via NextAuth.js and data integrity through strict Mongoose schemas.
- Uses ImageKit API for automated video transformations and thumbnail generation.
Flutist