Loading...
Q3 2025 Rankings
AI and ML engineering teams are moving at an unusual pace this quarter. We are tracking model infrastructure startups with commit velocities that have tripled in under three weeks.
Key Takeaway
In Q3 2025, 8 of 14 tracked ai & machine learning startups show positive engineering acceleration. harvard-edge leads with 748 commits over 14 days (+713% change). The dominant signal pattern is "Framework migration". Average sector commit velocity is 221 commits per 14-day window. These engineering momentum signals have historically preceded fundraise announcements by three to six weeks.
Data sourced from public GitHub activity. Read our methodology
Sorted by commit velocity change (14-day window, descending). Top 3 highlighted. Data last updated Q3 2025.
Get the free weekly engineering-acceleration rankings, or unlock the full Dashboard for real-time tracking, sector filters, and founder contact data. Beta pricing: €9.97/mo.
Want the signal while you research? Crunchbase/Wellfound badge or GitHub hover lookup — both free.
By signal type in this sector
In Q3 2025, we are tracking 14 ai & machine learning startups with measurable GitHub engineering signals. 8 of 14 show positive commit velocity growth. The most common signal type is "Framework migration", observed in 13 of the tracked companies. The average 14-day commit velocity across the sector is 221 commits, with harvard-edge leading at 748 commits (+713% change). These patterns have historically preceded fundraise announcements by three to six weeks.
harvard-edge leads the ai & machine learning sector in Q3 2025 with 748 commits over a 14-day window, representing a +713% change from the prior period. With 97 active contributors and 1 new repositories, harvard-edge is showing a "Engineering hiring burst" pattern — one of the more reliable leading indicators of a significant product milestone or fundraise.
Among the 14 ai & machine learning startups we track, US accounts for the highest concentration with 4 teams. Startups building AI/ML infrastructure, applications, and tools. Geographic distribution matters for investors because engineering talent clusters correlate with sector-specific domain expertise and proximity to early adopter customers.