AI/ML Developer · Python & Full-stack · Real-time Computer Vision
I’m a third-year CSE (AI-IBM) student who builds practical AI systems — from custom models and multi-agent pipelines to high-throughput real-time face recognition. I enjoy turning complex research ideas into reliable software and production-ready services.
Resume & contact details: Portfolio
- Fine-tuning LLMs and building agent/workflow orchestration with LangChain.
- Deployment & integration: FastAPI, Gradio, Next.js, Docker, Nginx.
- Real-time CV systems using YOLO (v11), face_recognition/OpenCV, and ChromaDB for vector retrieval.
- Building parameter-efficient models and training loops (PyTorch, mixed precision, optimization strategies).
-
Multi-Agents — Automated Product Development Orchestrated a 7-stage LangChain multi-agent pipeline (Product Manager, Engineer, Risk Analyst, Clarifier, etc.) to autonomously generate product blueprints. Deployed with FastAPI + Next.js + Docker. Repo:
https://github.com/Sambhav242005/Multi-Agents -
Real-Time AI Face Recognition System High-throughput face tracking using YOLOv11 + vector DB (ChromaDB), multi-process architecture (detection / encoding / rendering) for low latency and scalable identification. Repo:
https://github.com/Sambhav242005/AI-Face-Detect… -
Savant — Custom Universal Transformer Parameter-efficient recurrent transformer built in PyTorch with liquid gating and a logic-heavy training strategy to improve reasoning density.
-
Prime-Paradox (CampusCare) Multi-agent intelligent medical ecosystem using LangGraph & OpenRouter, Next.js frontend and FastAPI backend for real-time monitoring and doctor-patient workflows.
Areas of strength: Computer Vision (YOLO, tracking), Vector Databases (ChromaDB), LLM orchestration (LangChain), model fine-tuning, production deployment (Docker, FastAPI, Next.js), real-time systems.
- LinkedIn: Sambhav Surana
- Portfolio: https://sambhav.site
- Resume / CV (uploaded): Portfolio
- Need a realtime CV system demo? I can provide a Gradio/RT demo of YOLO + face recognition.
- Building agent orchestration? I can help design the LangChain stage flow and error-handling logic.
- Interested in model fine-tuning? I can help set up LoRA/PEFT pipelines and evaluation metrics.