Sahil Patel

About Sahil Patel

I am an AI/ML Engineer with around 2 Years of proven experience deploying production-grade computer vision, Physical AI and LLM systems from edge devices to cloud infrastructure. Specialized in real-time object detection pipelines, multi-object tracking, and RAG-based generative AI applications with proven expertise in reducing inference latency and scaling AI solutions across NVIDIA Jetson, AWS, and Azure platforms.Core Competencies:• Computer Vision & Edge AI: Real-time detection/tracking (YOLO, DeepSORT, TensorRT), embedded deployment on NVIDIA Jetson, CUDA optimization• Generative AI & LLMs: RAG system architecture, prompt engineering, LangChain agents, Azure AI Services integration• MLOps & Cloud: CI/CD pipelines, Docker containerization, model versioning (ClearML, MLFlow), AWS/Azure deployments• Production Systems: Multi-camera streaming (RTSP/GMSL), WebSocket integration, automated alerting, system monitoringKey Technical Achievements:• Engineered drowning detection system achieving 92.7% F1-score at 30 FPS across 4 concurrent camera streams• Optimized ReID tracking pipeline from baseline to 89.6% accuracy using YOLO + SAM embeddings• Deployed autonomous vehicle perception stack with traffic sign detection, lane keeping, and emergency braking• Built voice-enabled RAG coach using Azure TTS/STT with sub-200ms response latency• Architected blockchain-integrated document verification system processing documents/hour

Experience

1. Aiml Engineer

   ignatiuz

   Apr 2025 — Present · Indore, IN

   Project Engineer – AI, Machine Learning, and Computer VisionFocused on developing real-time detection and tracking systems that address critical challenges and contribute to life-saving applications- Engineered advanced detection and tracking pipelines for real-time inference in high-impact environments, delivering 92.68 percentage with 30+ fps in real time- Integrated Re-Identification (ReID) with YOLO-based models, achieving 89.6% accuracy in person/object tracking- Optimised tracking using SAM embeddings, improving performance from baseline to 94% in precision and efficiency metrics.Project aimed to be working on multiple deployment states such as cloud, edge devices and mobile applications. With integration in web and android apps.

Education

• IPS Academy Indore

  Bachelor of Technology - BTech, Artificial Inteliigence and Machine Learning