I explained how large language models learn and generate text, summarizing training, token-by-token generation, and refinement methods.

I prepared and cleaned pharmaceutical documents using Python in Google Colab, applied Pandas and image preprocessing, and produced cleaned data and enhanced images for OCR workflows.

I extracted text and structured field data from multi-page SDF PDFs using Python and produced a shareable deliverable.

I compared three OCR engines on a scanned pharmaceutical supplier PDF and documented their relative strengths and weaknesses.

I built and tested a complete RAG pipeline for document retrieval, selecting models, chunking, and hybrid retrieval strategies.

I ran three RAG configurations on an SDF sample and compared retrieved chunks, answers, and confidence to find the best retriever setup.

I built a page-level segmentation and metadata tagging pipeline to split bundled regulatory PDFs and label individual documents for routing.

I built a Gradio-based chatbot UI and integrated it with a local RAG pipeline to answer questions from uploaded PDFs.