That’s something we're working on completely
Kavin Stewart, Partner at Tribe Capital, on Reddit's 10x opportunity
Automating this work would turn Tribe from a partner driven investment shop into a repeatable data factory. The hard part in venture is not reading a pitch deck, it is getting messy company data into one clean format so revenue growth, retention, burn, and market position can be compared across thousands of startups. Tribe has built its identity around that normalization layer, and has since packaged it into Termina, an AI powered underwriting framework.
-
The current workflow is two jobs stitched together. First, collect raw operating data from founders, then map each company into the same schema so metrics are comparable. Stewart describes both the data cleanup and the writeup of underwriting reports, including benchmarks and strategic landscape analysis, as targets for automation.
-
This is consistent with Tribe's broader model. Other Tribe material describes a bottom up underwriting system built around product market fit, and later public materials show Termina becoming a named internal platform. That suggests the firm sees proprietary workflow software, not just partner judgment, as the core asset.
-
The closest analog is AI due diligence in adjacent markets like private equity, insurance, and legal work, where firms are using models to gather data, structure it, and draft first pass analysis. The common pattern is not replacing the final decision maker, but collapsing analyst hours spent on formatting, lookup, and synthesis.
The next step is venture firms behaving more like software companies, with smaller teams handling more deals because the machine does the intake, benchmarking, and first draft memo. If Tribe keeps compounding its internal dataset and automates the normalization layer, speed and consistency become a real moat, especially in markets where founders expect answers in days, not weeks.