Ankur Rathee

I’m an AI/ML-focused engineer pursuing a B.Tech in Computer Science with a track in Cloud Computing & Virtual Technologies, and I love taking models from idea to deployment. I build systems that are practical, measurable, and easier for teams to maintain.

In my AI Intern role at Omnie Solutions, I engineered end-to-end EDA pipelines in Python (Pandas, NumPy) across 5+ structured datasets, cutting preprocessing time. I also deployed 3 RESTful APIs using FastAPI to serve ML model predictions and processed data to internal analytics dashboards, handling 500+ daily requests.

Before that, as a Software Engineering Intern with DRDO (ITG Wing), I developed a secure SaaS platform that enabled browser-based MATLAB and Octave access for 50+ defence research engineers. I deployed and configured cloud-based lab infrastructure, reducing software access setup time by ~60% for the R&D team.

My projects reflect the same engineering mindset: RecallAI is a RAG-based document intelligence system using LangChain and FAISS, reaching ~90% answer accuracy on 10+ test documents and reducing hallucinations by 40% through dual-pass verification. Mentor Matrix delivers a real-time MERN mentorship platform with sub-200ms message latency, and Retainicta achieves 87% churn prediction accuracy via a Random Forest served through a Flask REST API with JWT and Google OAuth.