Prog.ai Recruits Software Engineers by Reading Their GitHub Commit History

01 /The Concept

Every company says it wants the best engineers. Prog.ai is betting they can actually find them – by reading the code.

The platform's core is a crawler that sweeps GitHub repositories, pulls commit code from individual contributors, and runs AI analysis to infer technical competencies. The robot works across well-known open-source projects, hobby repos, and even student repositories where learners submit coursework. Raw GitHub data isn't enough to build a full picture, so the system cross-references LinkedIn to fill in work history and contact details – using AI to resolve identity matches even when someone's GitHub handle and LinkedIn profile share only indirect signals like similar usernames or profile photos.

The result is a database of 60 million developers with inferred skills and contact information. Companies can search by competency, education, job title, and more, and the platform includes outreach tools for messaging candidates directly through LinkedIn or other identified channels.

Pricing isn't public – the platform is in beta – but the likely model is volume-based: companies pay per contact made with matched developers. Despite the early stage, the startup has already closed a $1 million seed round.

The founders signal that recruiting is just the first product. Advertising developer tools to a precisely segmented audience of 60 million engineers is another revenue line that the same database enables, with targeting granularity that general ad platforms can't replicate.

02 /Why It Matters

The most sought-after engineering candidates are passive – already employed, not looking, and therefore invisible to standard job boards. The only way to reach them is to identify them proactively and make the right offer at the right moment.

Using GitHub as a signal source for recruiting isn't new – technical recruiters have done it manually for years. What platforms like Prog.ai change is the scale and speed: what previously took a recruiter hours of searching can now surface in seconds across millions of profiles.

But finding thousands of technically qualified candidates doesn't solve the problem. A tighter signal matters more than a longer list. A [related review](/review/predlozhenija-nuzhno-delat-vovremja) covered Humanpredictions, which raised roughly $2 million and takes this further. Its platform combines GitHub and LinkedIn data with signals from over 40 specialized technical communities, then uses that behavioral layer to predict which engineers are psychologically ready for a career move – for example, because they've started engaging with technologies they don't use at their current job. That intent signal is more actionable than competency alone.

The next frontier for platforms like Prog.ai is exactly this: not just finding qualified people, but identifying who among them is already warming to the idea of switching – and understanding which specific offer framing would actually land. That level of precision wasn't achievable at scale before AI. It is now.

03 /Opportunities

Two structural forces are converging to make passive-candidate platforms more valuable. The chronic shortage of skilled technical talent isn't going away. At the same time, economic pressure is pushing companies to do more with leaner teams – which makes the quality of each hire matter more, not less.

The combination means companies will pay more, and compete harder, for the same small pool of exceptional engineers who aren't actively job hunting. Platforms that help surface and reach those candidates are well-positioned.

Prog.ai's approach – even at a relatively basic stage – was already enough to attract investor interest. A more behaviorally sophisticated version of the same infrastructure, closer to what Humanpredictions is building, represents the higher-value destination. The timing is right: the underlying AI capabilities that make this possible have only recently matured enough to deploy at meaningful scale.