Pfizer logoPfizer Externship via Extern
lorraine mureya portrait

Senior CS student building practical AI automation skills

Computer Science senior at Saint John's University focused on AI agent engineering and automation through a Wayfair n8n externship.

Work samples

Portfolio of coursework and work from the Wayfair n8n AI Agent Engineering externship, showing practical experiments in AI automation and agent workflows.

  • Pfizer Advanced: AI-Powered Document Insights & Data Extraction Externship
    Pfizer Advanced: AI-Powered Document Insights & Data Extraction Externship
    Pfizer logo

    Pfizer · ✅ Verified by Extern · ⏱️ In progress

    Pfizer Advanced: AI-Powered Document Insights & Data Extraction Externship

    Prototype AI-powered document intelligence with Pfizer—using OCR, LLMs, and RAG to automate real enterprise PDF workflows and build a standout portfolio project.

    AI & MLPythonDocument IntelligencePresentation Skills

About me

Computer Science senior at Saint John's University focused on AI agent engineering and automation through a Wayfair n8n externship.

I am Lorraine Mureya, a senior Computer Science student at Saint John's University pursuing a bachelor's degree. I have limited formal work experience and am completing the Wayfair n8n AI Agent Engineering externship as part of my learning journey.

Externships

Pfizer Advanced: AI-Powered Document Insights & Data Extraction Externship

In progress

Skills

PythonAI agent conceptsn8n automationBasic algorithmsProblem solving

✅ Verified by Extern · ⏱️ In progress

Pfizer Advanced: AI-Powered Document Insights & Data Extraction Externship

Prototype AI-powered document intelligence with Pfizer—using OCR, LLMs, and RAG to automate real enterprise PDF workflows and build a standout portfolio project.

AI & MLPythonDocument IntelligencePresentation Skills

Overview

The externship prototyped AI-powered document intelligence for enterprise PDF workflows using OCR, LLMs, and retrieval-augmented generation. The work surveyed foundational LLM concepts, implemented approaches for extracting and indexing PDF content, and demonstrated end-to-end document query and summarization prototypes.

Pfizer Advanced: AI-Powered Document Insights & Data Extraction Externship

What I've accomplished

I documented LLM training processes, explained transformer context mechanisms, and showed how model outputs are generated via next-word probability selection.

Project breakdown

The work summarized how LLMs were trained on large text corpora to learn language patterns, described the transformer mechanism for context-aware token prediction, and explained that outputs are produced by next-word probability selection.

The module showed preparing data for AI: I cleaned tabular data with Pandas, flattened nested JSON into a normalized DataFrame, performed text standardization steps, and applied image denoising, contrast and thresholding to improve OCR input.

The project processed multi-page SDF PDFs with a Colab Python notebook that extracted text and bounding boxes, applied regex and anchor-phrase logic, and produced structured field-level data for dates and vendors.

I ran Tesseract, EasyOCR, and PaddleOCR on a scanned supplier PDF, noted installation and preprocessing issues, compared outputs for tables, lot numbers, and dates, and produced a report that found PaddleOCR gave the most accurate, layout-preserving results.

Google Docs
Access

The module required building and optimizing a RAG pipeline. I ingested documents, implemented chunking and embeddings, tuned retrieval settings, and ran queries. The submission produced a Colab pipeline and a Google Doc of retrieval test results.

Google Docs
Access

I tested three top_k configurations on an SDF sample, retrieved and compared chunks, and found that retrieving 5 chunks gave the most efficient and least redundant results; larger top_k values added repeated or unrelated text and did not improve the answer.

Google Docs
Access
View all work