I'm an AI Engineer with 2 years of hands-on experience building and shipping production-grade AI systems — from multi-agent frameworks and LLM-powered decision engines to large-scale computer vision pipelines. I've delivered end-to-end AI solutions for clients across the USA, Saudi Arabia, Germany, and the Middle East, spanning e-commerce, government, manufacturing, and automotive sectors.

🏅 Ranked Top 0.1% (64th of 20,000+ teams) in the Amazon ML Challenge 2024, a 5X Hackathon Winner, and certified as a Microsoft Azure Data Scientist Associate and Databricks Certified Data Engineer Associate. I thrive in fast-moving environments where owning problems end-to-end and shipping real impact matters.

• Multi-agent AI frameworks (CrewAI, LangGraph, MCP) for enterprise decision-making and evaluation systems
• Large-scale computer vision pipelines for surveillance, tracking, and workforce productivity analytics
• On-premise deployment of open-source LLMs (Qwen2.5 VL, Qwen3) for security- and compliance-sensitive environments
• Benchmarking state-of-the-art ASR and TTS models for high-throughput, low-latency voice AI

• Designing and shipping multi-agent AI systems with CrewAI, LangGraph, and MCP (Model Context Protocol)
• Building and fine-tuning computer vision pipelines (YOLO, RF-DETR) with tracking/ReID (BoT-SORT, ByteTrack, DeepSORT)
• Architecting production RAG, LLM-powered decision engines, and agentic workflows at enterprise scale
• On-prem deployment of open-source LLMs, ASR, and TTS models for compliance-heavy environments
• MLOps practices for continuous integration and deployment of ML systems

Feel free to reach out at neeldevenshah.ai@gmail.com

"I spend my days training AI agents to make smart decisions — and my evenings still can't decide what to eat for dinner."

• AI & ML: Deep Learning, Computer Vision, NLP, Reinforcement Learning
• Generative AI: LLMs, LoRA, RAG, Reranking, Guardrails, Diffusion Models, TTS/STT
• Agentic Systems: CrewAI, LangGraph, LangChain, MCP (Model Context Protocol)
• MLOps: Git, Docker, Kubernetes, Kafka, MLflow, W&B, Apache Airflow, CI/CD, VLLM, Unsloth

• 🏅 Top 0.1% (64th of 20,000+ teams) — Amazon ML Challenge 2024
• 🏅 Top 0.3% (All-India Rank 80 of 23,000+ teams) — Amazon ML Challenge 2025 (Price Prediction)
• 🏆 5x AI Hackathon Winner
• 🥇 1st place, IndoML Datathon 2025 – Phase 1 (Track 1) and Phase 2 (both tracks)
• 🥈 2nd place, IndoML Datathon 2024 Phase 1 (350+ teams)
• 🏅 Winner, The Maverick Effect AI Challenge
• 🎖️ Top 11% (23rd of 196 teams) — Meesho Visual Taxonomy Data Challenge, CODS-COMAD
• 🎓 Microsoft Certified: Azure Data Scientist Associate
• 🎓 Databricks Certified Data Engineer Associate
• 🏛️ Finalist, AI Hackathon 2024 at IIM Ahmedabad

Let's connect and chat! Open to exciting AI collaborations.