Hemant Dadhich

I am an AI/ML engineer with hands-on experience building production-ready computer vision systems and LLM-powered agents. I have delivered end-to-end solutions that span model training, optimization for edge devices, and cloud-based retrieval/ingestion pipelines.

At Abstrabit Technologies I designed AI voice and chat sales agents, a RAG evaluation system using AWS Bedrock and Gemini, and a MedPrompt-based healthcare triage that reduced unnecessary ER visits by 25–30% while maintaining HIPAA compliance.

At Rawattech I built meter-reading and video-retrieval computer vision solutions, optimized models for Google Coral Edge TPU through quantization and TFLite conversion, and performed large-scale data cleaning and rigorous model evaluations to improve accuracy.

I enjoy deploying lightweight models on edge devices (Raspberry Pi, Coral), integrating LLMs and retrieval systems, and optimizing real-time inference pipelines to deliver measurable business impact.