vivek User

I’m a results-driven ML Engineer and AI practitioner with hands-on experience building and deploying end-to-end AI systems. I focus on shipping production-ready solutions—model training, embedding pipelines, real-time inference APIs, and containerized deployments.

In my Data Science internship, I delivered 5+ ML projects (classification and regression) using Python and Scikit-learn, working with 50,000+ record datasets through full data cleaning, feature engineering, and EDA with Pandas and NumPy. I improved classification performance up to 90% accuracy (from a 74% baseline) via cross-validation and hyperparameter search, and I reduced client reporting time by 30% through dashboarding and evaluation reports.

My projects reflect how I architect GenAI products: I built a 5-agent LangGraph career guidance + resume intelligence platform with hybrid RAG and an ATS scoring engine, exposing models as real-time FastAPI endpoints and keeping the backend non-blocking with async Redis + Celery queues. I also built a semantic memory assistant with sub-200ms query latency using FAISS and ONNX Runtime, and an AI disease prediction + priority scheduling system that serves structured JSON predictions and triage via Docker Compose–containerized FastAPI APIs.