SAMEER.AI
AI Engineer — RAG & Agentic Systems
M.Sc. Data Science candidate at IIIT Lucknow, currently a Data Analyst intern at Obel Systems. I build retrieval-augmented and agentic AI systems — grounded RAG pipelines, MCP-connected agents, and fine-tuned deep learning models — that stay honest about what they actually know.
About me Get in touchRemote
Data Analyst Intern
Lucknow, India
Teaching Assistant
India
Flight Cadet
121+
Automated tests passing across shipped projects
90%
Peak model accuracy (fine-tuned ResNet50 X-ray classifier)
4
AI/agentic systems designed and shipped
3+
Hackathon & challenge builds (Google Cloud, Redrob, and more)
A proactive commute-planning agent for Lucknow. It simulates future departure windows, fuses traffic, weather, festival, and event risk into a deterministic score, and uses MongoDB MCP as agent memory over commute history. Built for the Google Cloud Rapid Agent Hackathon, MongoDB Partner Track.
An end-to-end RAG system for document-grounded Q&A over typed, scanned, and handwritten PDF notes. OCR fallback covers messy scans; strict grounded prompting keeps answers faithful to the uploaded material instead of hallucinating.
A hybrid candidate-ranking engine built for the Redrob Data & AI Challenge. Scores candidates on semantic match, career evidence, and behavioral signals instead of keyword matching, producing an auditable top-100 shortlist.
A pneumonia classifier fine-tuned from ResNet50 with transfer learning, differential learning rates, and data augmentation, with Grad-CAM visualizations for interpretability.