Skill Extractor¶

Skill Extractor is a research tool that uses local LLMs (Llama) to parse project description PDFs and automatically identify and rate the skills required to complete each project.

How It Works¶

Given a project description PDF (e.g., a capstone project specification), Skill Extractor:

1. Parses the PDF to extract the project description text
2. Feeds the text to a local Llama model with a structured prompt
3. Returns a JSON-formatted list of required skills with necessity ratings

The output feeds into downstream tools like OptiTeam for team formation and SkillNet for skill graph construction.

Status¶

This was an early proof-of-concept that predates OptiTeam. Its core idea — using LLMs to extract structured skill data from unstructured text — has been absorbed into the broader SkillNet and OptiTeam pipelines.

Related Research¶

• Skill Networks — The research framework that absorbed this project's extraction pipeline.
• Algorithmic Team Formation — Downstream consumer of extracted skill data.