DeVaris Brown
Cofounder
Cofounder and CEO of Meroxa, the real-time data infrastructure company. Previously a product leader at Twitter, Heroku, VSCO, and Zendesk. Angel investor and advisor to dozens of startups.
Venture studio · Multi-strategy fund
We build and back the AI companies that will still matter in thirty years.
A venture studio and multi-strategy fund from DeVaris Brown and Van Jones. We start companies, back founders, and buy businesses where AI permanently changes the economics.
01 / Partners
An operator and an investor. One firm.
Why Third Letter. DeVaris and LaVandez. The third letter of each name is a capital V.
LaVandez “Van” Jones
Cofounder
Venture investor. Deal Lead at Wellington Access Ventures, the early-stage platform of Wellington Management, and previously a Partner at Drive Capital focused on enterprise SaaS and data infrastructure. Cofounded Hello Tractor, an agricultural equipment platform operating across Africa. Chicago Booth MBA.
02 / Thesis
AI is not a sector. It is a change in what a durable company costs to build.
Origination
We meet companies before the round exists.
Both partners build in public. Deal flow comes from founders, operators, and communities we already work alongside.
Structure
The vehicle is chosen last.
Studio, venture, control buyout, or public equity. Whichever fits the company.
Duration
Legacy is a time horizon.
We underwrite to decades, hold through cycles, and measure ourselves on what survives.
03 / Model
One thesis. Four instruments.
The studio builds the tooling. Venture backs the companies using it. Private equity applies it to businesses that already make money. Public markets capture the re-rating.
- Studio
Companies we originate, staff, and spin out.
We build from problems we have lived as operators, supply the first team, and hold founding equity. Kno is the first.
- Venture
Early rounds in AI-native companies.
We back founders where a technical wedge can compound into a durable moat, and bring operating experience to the table.
- Private Equity
Profitable businesses we can re-tool.
Services and software companies whose operating core can be rebuilt with the studio’s tooling. Value comes from margin, not multiple.
- Public Markets
Long-term positions in listed companies.
Incumbents where AI adoption changes unit economics before it changes the multiple.
04 / From the studio
Kno
Studio company 01Know which data actually makes your AI better.
Open-source agent data evaluation. Kno measures which documents, examples, policies, and conversations improve an AI agent, and by how much, then ranks each by impact and cost.
- Open source
- Apache-2.0
- Single Go binary
- v0.1
| Asset | Impact | Decision |
|---|---|---|
| new_refund_policy.md | +18% | keep → knowledge_base |
| example_42.json | +7% | keep → context |
| example_91.json | +1% | reject |
| old_refund_policy.md | −9% | reject → harmful |