I summarized how large language models work and the training stages that shape their behaviour.

I prepared and cleaned pharmaceutical document data using Python in a Google Colab notebook, including data-structure exercises and preprocessing for OCR.

I extracted and structured text from multi-page SDF PDFs using Python, then provided the resulting export file.

I compared OCR engines and produced a reproducible Colab that extracted and evaluated text from a scanned pharmaceutical PDF.

I built a simple chatbot using LlamaIndex in a Colab notebook and demonstrated retrieval-backed responses.

I compared open-source LLMs and retrieval settings in a RAG pipeline and recorded results from multiple Colab experiments.

I built a page-level segmentation and metadata extraction pipeline for bundled regulatory PDFs, producing tagged document boundaries and metadata for downstream routing.

I built a Gradio chatbot UI and connected it to a local RAG pipeline for document Q&A.

I documented Fulfillment Center structure and common frontline roles, mapping how tasks and handoffs relate to customer-facing metrics and the Learning Ambassador function.

I analyzed web-scraped employee reviews to identify which workers showed early-tenure attrition risk, extracting red-flag signals and patterns that correlate with operational disruption.