Seeking AI/ML research internships and software engineering internships. Strong in full-stack engineering. Moving deeper into ML systems.
B.Tech CSE @ KIIT, Bhubaneswar (Graduating 2027)
LinkedIn · Email
- PyTorch LSTM Autoencoder trained on the MIT-BIH Arrhythmia Database — unsupervised, trained only on normal heartbeats
- Anomalies flagged via reconstruction error thresholding; follows AAMI EC57 patient-aware train/test split to prevent data leakage
- Stack: Python · PyTorch · wfdb · NumPy · scikit-learn
- Real-time BFS, DFS, and A* on procedurally generated mazes using a 4-component decoupled architecture
- A* with Manhattan heuristic expands ~60% fewer nodes than BFS while maintaining optimal path length
- Stack: Python · Pygame · Ursina 3D
- AI-powered fitness platform using Groq AI to generate personalized meal plans via server-side cron delivery
- Multi-tenant data isolation enforced via Supabase Row Level Security (RLS) and JWTs
- Stack: Next.js 15 · TypeScript · Supabase · PostgreSQL · Groq AI
- Real-time multiplayer drawing game — 10+ concurrent players, sub-50ms WebSocket broadcast latency
- Authoritative server-side state machine handles room lifecycle, turn logic, and score arbitration
- Stack: React · Node.js · Express · Socket.io
ML/AI | Python · PyTorch · NumPy · scikit-learn · LSTM · Autoencoders
Algorithms | A* · BFS · DFS · Dynamic Programming · Graph Search
Web | React · Next.js 15 · Node.js · TypeScript · PostgreSQL · MongoDB
Tools | Git · Docker · Vercel
Open to AI/ML research internships and software engineering internships — krishnakumardey452@gmail.com