Workflow Driven Private Market Liquidity
Alex Johnson, co-founder & CEO of Velvet, on vertical AI for venture capital
The real wedge is not better matching logic, it is owning the daily workflow data that reveals who wants what before anyone formally comes to market. Velvet is trying to sit upstream of the transaction, where investors read decks, draft memos, track theses, and update CRM records. If that behavior lives inside one private workspace, an agent can infer likely buyers, sellers, co investors, and secondary interest without exposing raw firm data.
-
This works only if the product is used every day. The interview frames the terminal model clearly, research, data, participants, and workflow in one place. That is why the analyst product comes before the liquidity layer. The software captures the signals, then the network can suggest matches through double opt in.
-
The closest precedent is private market infrastructure players buying data exhaust. J.P. Morgan bought Aumni for venture analytics and document data, and BlackRock bought Preqin for private markets data and relationships. The pattern says proprietary workflow and market data can become more valuable than a point solution for a single transaction.
-
It also explains the emphasis on privacy and conflict controls. Carta had the ownership ledger but struggled to turn that into a trusted exchange workflow. Velvet's model is to keep each fund in a sandboxed environment, then let agents suggest connections only when both sides choose to engage, which is closer to liquidity discovery than price discovery.
If this model works, private market liquidity will shift from broker led searching to software led discovery. The winners will be the platforms that see investor behavior earliest, turn it into useful intent signals, and route those signals into secondaries, syndicates, and co investments without breaking trust. That is how an AI analyst can expand into a market access layer.