Nicholas Parmigiano Data Analyst
Statistics · UF ’26

Open to Data Analyst, Legal Ops & Compliance roles: South Florida or remote

Work / Independent / 2025

Legal Document Analyzer In progress

Started Jan 2025 · working MVP, used on real case files

Upload a legal document and get back its type plus a structured timeline of events. I built it because I was the paralegal doing that review by hand.

Headline
MVPused on real legal work
Stack
JavaScriptNode.jsReactRule-based extractionOpenAI API (v1)
INCIDENT TREATMENT DEMAND FILING MEDIATION
Unstructured filing → extracted timeline · illustrative sketch, not the source data

The problem

At high case volume, reading every medical record and filing to rebuild “what happened when” is slow and easy to get wrong. I lived that problem at Marz Law and Full Focus Law.

What it does today

The MVP takes an uploaded document, classifies it (contract, motion, deposition, medical record…) and extracts a dated timeline of events. I’ve used it on my own real case work.

What broke, and the fix

Version one sent documents to the ChatGPT API. When that AI component stopped working, I didn’t wait on it. I hard-coded the classification and extraction logic so the tool kept doing its job. Rules are less flexible than a model, but they’re predictable and they never make things up, which matters with legal documents.

What’s next

Public repo with a README and a demo, then deciding where a model really earns its place on top of the rules. This page will update as it ships. Follow along in the journal.

Ask AI about me

“Tell me about Nicholas Parmigiano based on https://nickp-site.pages.dev. Summarize who they are, what they do, and how to get in touch.”

ChatGPT Claude Perplexity Gemini