GradPipe

GradPipe parses repositories at the abstract syntax tree level, which means it looks at how code is structured rather than how a profile is worded. A student who refactored a messy project into clean, reusable modules that other people can actually read shows up differently from one who only forked a template and tweaked a few values. For a hiring manager, that's a signal a keyword filter simply can't produce.

This is the core idea behind the whole product, and everything else supports it. GradPipe centers on GitHub proof-of-work over self-reported claims. It checks what a candidate has actually built and committed. A portfolio carries more weight than a skills list. Simple as that.

GradPipe is aimed at the "undiscovered" part of the market: students whose work is strong but whose networks aren't. That focus helps teams reach candidates who never make it onto a recruiter's radar through traditional channels. It also means the pool skews toward people early in their careers.

Rather than tagging candidates with skills they simply claim on a form, the platform infers strengths from the repositories they maintain over time. A backend-heavy GitHub tells a different story than a frontend one that leans on layout and interaction work. Recruiters get a far more concrete starting point for a conversation than a list of self-assigned labels.

Results arrive ordered by how well each profile matches the role, so a hiring team starts with the strongest matches instead of a flat list. That ordering is the difference between a search tool and a shortlist tool.

Hiring teams describe the role and requirements, then let the platform surface relevant student profiles. The workflow assumes you already know roughly what you need and want help finding it faster, rather than help deciding what to hire for in the first place. That's a real assumption. Not every team has a clear role spec on day one.