AI Externship @ Outamation

Designed and implemented end-to-end AI pipelines for mortgage document automation using Python, integrating NLP, OCR (Tesseract, Label Studio), PyMuPDF, and Computer Vision to extract and classify structured data from unstructured PDFs, significantly improving downstream model accuracy.

 

Engineered and optimized Retrieval-Augmented Generation (RAG) systems using LlamaIndex, Llama, and Deepseek, leveraging Python, LangChain, and vector-based search frameworks to enhance document retrieval and QA. Authored in-depth technical reports on model benchmarking and deployment strategies for scalable mortgage AI solutions.