Sumit Kumar

I’m an AI Engineer and Machine Learning Engineer with 7+ years of experience designing and deploying production-scale machine learning and agentic AI systems across healthcare, enterprise platforms, and global trade compliance. I’m currently leading the development of an AI-powered Harmonized System (HS) code classification platform using hybrid deep learning and agentic LLM architectures.

I build end-to-end ML pipelines—from data ingestion and feature engineering to model training, fine-tuning transformer models, scalable deployment, evaluation frameworks, and observability pipelines. In my current role, I’ve designed agentic workflows with LangGraph and LangChain, implemented RAG using Qdrant, exposed classification APIs with FastAPI, and fine-tuned DistilBERT with PyTorch/HuggingFace on AWS SageMaker, deploying low-latency inference with NVIDIA Triton while monitoring drift, quality, and latency.