Gopi Sundaramurthy came to venture capital from computational biology, starting at Cellworks and later working as the first data scientist at IBM Watson Health, and he built that background into Unity, a system he spent five years developing that ranks startup teams using hundreds of data points and is now used by both VCs and Wilson Sonsini. At Ensemble, he keeps the final invest decision in human hands even as models help with sourcing, insists that good data needs matching firm-wide processes to be useful, and keeps his team disciplined by limiting itself to about three projects at a time and shutting down ones that aren't working.
About their Fund
Ensemble VC is unusually software-native, with about half its team made up of data scientists and engineers who spent five years building Unity, a platform that ranks startup teams using hundreds of signals and is now also used by outside firms like Draper Associates, Boost, Overmatch, and Wilson Sonsini; the fund invests based on what teams know rather than who they know, puts over 50% of its management fee into this data platform, and reaches out proactively to a small number of high-conviction founders each year, offering a 360 review of leadership teams as companies grow. Based in Texas rather than Silicon Valley, the founding team combines classic venture backgrounds from Kauffman Fellows, Bain Capital Ventures, Norwest, and Correlation Ventures with data science experience from IBM Watson, and it uses talent-flow analysis to spot emerging categories early, as seen in investments like Blueprint, Loti, Saronic, CHAOS, and Manifest, while also aiming to identify people who could become founders and encourage them to start companies.
















.jpeg)







.jpeg)











































































.png)
























































































































.jpeg)









.jpeg)




















































































