I summarized how large language models generate text by predicting tokens from learned probabilities.

I prepared and processed pharmaceutical documents using Python in Google Colab, applying data cleaning, JSON handling, text standardization, and image preprocessing to improve OCR readiness.

I extracted and structured text from multi-page SDF PDFs using Python tools, then identified key fields like vendor, document type, lot, and manufacture/expiration dates.

I compared Tesseract, PaddleOCR, and EasyOCR on a scanned pharmaceutical PDF and reported which engine performed best for layout-aware extraction.

I built and tested a complete RAG pipeline for document retrieval using LlamaIndex, selecting embeddings, chunking, and retrieval methods and validating answers against source documents.

I tested three RAG configurations on an SDF document, compared retrieved chunks and answer confidence, and selected the configuration that produced the clearest, most relevant explanation.

I built a page-level metadata extractor and boundary detector that labeled each page and grouped pages into document IDs by type.

I built a Gradio-based RAG chatbot UI and backend that extracted text from uploaded PDFs, indexed chunks with embeddings in FAISS, and returned context-grounded answers with source citations.