Velora Custom LM Trainer
A from-scratch PyTorch trainer for a compact LLaMA-style language model with resumable training and guided structured generation.
Hi, I'm Sambhav Surana
I build AI agents, computer vision systems, and full-stack ML products that turn research ideas into working software.
A from-scratch PyTorch trainer for a compact LLaMA-style language model with resumable training and guided structured generation.
A privacy-first, locally-run AI fitness coach combining YOLOv11 pose estimation with a vision-language model for real-time rep counting, form analysis, and AI coaching.
A full-stack micro-city MVP with a FastAPI simulation engine, React/TypeScript dashboard, interactive canvas map, and a local Q-learning policy agent — no external AI APIs required.
A local, biologically-inspired agentic memory system that organizes information into hierarchical concept groups and cells, going beyond flat RAG with three-stage hybrid search, entity routing, and LLM-driven consolidation.
A Python tower-defense game with 10 escalating waves, 50 dynamic path combinations, tower upgrades, economy, and a live GitHub Pages build.
A client-side AI face tracking and recognition application running YOLOv11 face detection and EdgeFace embeddings directly in the browser via WebGPU and ONNX Runtime.
Coursework: machine learning, NLP, time series, statistics, and full-stack development
Built AI timetable-generation work and ran Python, NumPy, Pandas, and Scikit-learn workshops
Velora transformer training, AI fitness coaching, multi-agent workflows, YOLO face recognition, and simulation systems
NPTEL Online Certification, IIT Kharagpur
36-hour hackathon focused on rapid prototyping, teamwork, and product problem-solving
How to combine centroid search, parallel per-group HNSW queries, and entity routing into a recall-safe hybrid search pipeline.
Designing a consolidation pass that merges similar concept groups using both vector similarity and LLM-based contradiction analysis.
How to structure persistent agent memory using concept groups and detailed cell retrieval in ChromaDB collections.