Use Perplexity and ChatGPT to locate a healthcare dataset tied to your research question, then build a lightweight Replit app to inspect the data, surface pain points, and shortlist the angles you want to pursue.

In this project, you will cover the basics of data analytics, including its importance and real-world applications in healthcare. You will learn about social determinants of health and healthcare-related data to understand their sources, types, and relevance in public health decision-making.

This project will introduce you to the fundamentals of data collection and pre-processing for analysis. You will gather relevant data from public sources and use LLMs to explore possible analyses. You will then clean and pre-process the data to ensure it is ready for analysis.

In this project, you will develop skills in exploratory data analysis (EDA) to extract insights from the relevant datasets. You will learn key statistical concepts, apply Python-based analysis techniques, and create visualizations that effectively communicate findings. Through hands-on activities in Google Colab, you will explore trends, relationships, and distributions in the data.

Before AI can process documents, it must first understand them. In this project, you’ll explore Machine Learning, Deep Learning, LLMs, and NLP to see how AI reads and processes text. You’ll also dive into Computer Vision and OCR, learning how AI extracts information from scanned documents. Since this externship focuses on mortgage automation, you’ll study the mortgage industry, its challenges, and why accurate document handling is essential. Finally, you’ll work with a real mortgage "blob" file, analyzing how AI models handle structured and unstructured data. By the end, you’ll have a strong AI foundation and a clear understanding of how AI improves document processing in the mortgage industry.

AI models require clean, structured data to function effectively. In this project, you’ll learn how to use Python for data processing, cleaning, and organizing mortgage documents to prepare them for AI automation. You’ll begin by exploring Google Colab and learning Python basics. Then, you’ll dive into data preparation techniques to clean and standardize information. Finally, you’ll work with image processing to enhance OCR (Optical Character Recognition) accuracy. By the end of this project, you’ll be able to prepare mortgage data for AI models, ensuring accurate and efficient automation.

Before AI can make sense of documents, it needs clean, structured data. In this project, you’ll learn how to turn both digital and scanned PDFs into machine-readable formats using Python. You’ll explore tools like PyMuPDF and pdfplumber to extract text from multi-page mortgage files. Then, you’ll apply field heuristics—like regex patterns, anchor phrases, and layout clues—to identify key fields such as borrower name, salary, and loan amount.

OCR (Optical Character Recognition) is at the heart of document intelligence. In this project, you’ll get hands-on with three OCR engines— Tesseract, PaddleOCR, and EasyOCR —to see how they perform on complex mortgage documents. You’ll start by learning how to clean and extract text from scanned PDFs using Tesseract. Then, you’ll explore layout-aware tools that better preserve formatting and structure. Finally, you’ll run comparisons and recommend the most effective tool for real-world automation. By the end, you’ll be able to confidently select and justify OCR tools based on both technical and business needs.

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.

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 mortgage documents and the ability to justify your design decisions based on retrieval accuracy and model efficiency.

Mortgage files often come as massive, unstructured "blobs" that include dozens of document types in a single PDF. In this project, you’ll design a system to split unstructured blobs into individual documents, classify each one, and route them to the appropriate extraction logic. You’ll use layout-based clues and rule-based methods—or even simple machine learning—to separate and identify documents like pay slips, IDs, and contracts. Then, you’ll build a routing function to apply the right extraction pipeline for each document type. By the end, you’ll have created a modular, intelligent system that mirrors real-world document workflows in mortgage automation.

In this module, you'll learn how to turn your local RAG pipeline into an interactive chatbot using Gradio. Instead of running queries in a notebook cell, users can ask questions through a clean web-based UI — just like ChatGPT. You'll split your code into two functions: one that builds the Gradio interface, and one that handles user prompts by sending them through your retrieval engine and LLM. This structure not only helps modularize your pipeline, but also makes it easy to plug in different models — including Gemini, Mistral, Phi-2, or any LLM of your choice. By the end, you'll have a working local chatbot that can answer questions about a PDF document.

I documented my initial data exploration for the CIYJ dataset and noted where key metadata and structure lived.

I mapped the five CIYJ files, documented each file's unit of analysis, row count, required preprocessing, and the standalone question each file could answer.