At Sword, I'm building an autonomous agent that scans learner profiles, surfaces skill gaps, and personalizes training recommendations. It can also update plans, assign modules, and manage learner records through the organization’s CMS.
At Xebia, I built a compliance-grade Medical AI pipeline that generated Rs. 2+ crores in revenue in 3 months, extracting information from 10,000+ documents with 95%+ accuracy. I also developed a prompt version control framework with 200+ prompts and an enterprise RAG repository indexing 8,000+ documents.
I engineered XParser, a document intelligence platform sold for $100K+ per deployment to healthcare clients in the USA and Europe. Its OCR, computer vision, and LLM pipeline extracts text, tables, and images from unstructured documents.
My Xebia work also included banking AI, where I built transaction categorization that achieved 92% accuracy across 500K+ monthly transactions and implemented a decision-support engine that increased loan uptake by 25%. I’ve also worked on fraud detection, MLOps, and projects including an open-source document parser and an AI nutrition app.
