Project 1: How AI Reads Pharmaceutical Documents

I examined how AI comprehends pharmaceutical documents by studying Machine Learning, Deep Learning, LLMs, and NLP. I investigated the role of Computer Vision and OCR in extracting data from scanned files, ultimately analyzing a real pharmaceutical document to understand how AI models process both

Project 3: Data Extraction from Documents Using Python

I transformed digital PDFs into machine-readable formats using Python, exploring tools like PyMuPDF and pdfplumber to extract text from pharmaceutical SDFs. I applied field heuristics such as regex patterns and layout clues to identify key fields, establishing a foundation for Pfizer's compliance

Project 5: Introduction to Retrieval-Augmented Generation (RAG)

AI models can process vast amounts of text, but finding the right information quickly is a challenge. In this project, you’ll learn how to build a Retrieval-Augmented Generation (RAG) pipeline using LlamaIndex to help AI retrieve relevant data from large document sets.

Project 7: Blob Processing, Classification, and Routing

I designed a modular system to process large regulatory PDFs containing multiple document types. This involved splitting these files into individual documents, classifying each one, and routing them to the appropriate extraction pipeline using layout-based clues and rule-based methods.

Project 9: Final Integration, Testing, and Evaluation

I integrated all components into a complete, end-to-end document intelligence system. I ran the pipeline on a full pharmaceutical SDF document, rigorously testing its performance and creating a clear, professional demo that showcased the system's real-world potential.

Project 2: Learn to Work with Python for AI-Powered Document Processing

I utilized Python for data processing and cleaning of pharmaceutical documents, ensuring they were well-structured for AI automation. I engaged with Google Colab to learn Python basics, employed data preparation techniques to standardize information, and enhanced OCR accuracy through image

Project 4: Advanced OCR Comparison and Layout-Aware Extraction

I engaged with three OCR engines—Tesseract, PaddleOCR, and EasyOCR—to evaluate their performance on scanned pharmaceutical supplier documentation. I learned to clean and extract text using Tesseract, explored layout-aware tools for format preservation, and made recommendations on the most effective

Project 6: Advanced RAG and Open-Source Experiments

In this project, you’ll take your RAG pipeline to the next level by improving precision and experimenting with open-source LLMs. First, you’ll explore advanced chunking techniques—like overlapping and tuning—to improve document segmentation. Then, you’ll implement metadata-based filtering so your system retrieves more contextually relevant answers. You’ll also evaluate the performance of open-source models like Mistral and Phi-2 compared to proprietary options. By the end, you’ll have a fine-tuned, high-performing RAG pipeline optimized for real-world pharmaceutical documents and the ability to justify your design decisions based on retrieval accuracy and model efficiency.

Project 8: Gradio Chatbot with RAG Integration

I developed an interactive chatbot using Gradio that integrated my local RAG pipeline. This allowed users to ask questions through a clean web-based UI. I structured the code into two functions for building the Gradio interface and processing user prompts, enabling easy integration of various LLMs.