Operator-led diligence
Deal-by-deal capital
Awaiting quorum
Technical conviction is a collective judgment.
The challenge
Access is the bottleneck.
VCs
Expertise and early access, on a deadline.
Paid networks are noisy and conflicted.
Founders
Few experts to sharpen the story. Fewer buyers.
The value proposition goes untested.
Operators
Deep judgment, no path into deals.
No on-ramp from operating into investing.
The solution
One neutral platform.
- Intelligence layer
- Structured diligence & expert access Technical review run by people who have built and operated the systems in question — not a survey of opinions.
- Capital layer
- Co-investment, deal by deal One vehicle per investment. We never lead, and we don't compete with the round.
Experts power the diligence. Capital follows the conviction.
Operating model
We complement, never compete.
- Never lead
- Co-investment only We come in alongside the lead, on the round's terms.
- Self-funding
- Subscriptions cover the platform The diligence work sustains itself, so the capital decisions stay unconflicted.
- Bidirectional
- Deal flow runs both ways Rounds come in for diligence; technical opportunities flow back out as leads.
The people
Three tiers, rewarded for value created.
| Tier 1 · General partnersCore operators | Tier 2 · Domain expertsEarned ownership | Tier 3 · Buyer networkAdoption signal | |
|---|---|---|---|
| Who | Full-time operators — senior product, engineering and AI leaders. | Pre-filtered experts across our nine domains — active, not full-time. | Decision-makers inside the enterprises that adopt this technology. |
| Role | Own decisions, lead diligence, back founders. | Source deals, run technical diligence. | Validate demand — what enterprises actually buy. |
| Reward | Carry, full deal flow, management fee. | Co-investment, contributor carry pool. | Co-investment on deals they validate. |
| Commitment | All-in — lead the work, don't advise. | Consistent contribution where you're strongest. | A candid answer on what you'd buy. |
Expertise areas
Nine domains.
Portfolio
Where we've already invested.
Foundations
- Baseten
- LanceDB
- WarpStream
- CedarDB
- Chalk AI
- Nimbleway
- VillageSQL
- Solid
Building with AI
- Dosu
- Unblocked
- Extend AI
- Zep AI
- Vellum
- Patronus AI
- Openlayer
- Eppo
- Sundial
- HiddenWeights
- G5
Enterprise readiness
- Runlayer
- Rox
- Trent AI
Twenty-two companies across the stack, with further positions not yet announced.
Team
Small, senior, AI-native.
Philip Zelitchenko
Founding partner
Two decades leading data science, AI platform, and data foundation teams at scale.
Ioannis Papapanagiotou
Founding partner
Distributed systems and data platforms; dual PhD, former Purdue faculty.
Ippokratis Pandis
Founding partner
Architect of petabyte-scale data systems; VP/Distinguished Engineer on Redshift; CMU PhD.
Ali Dasdan
Founding partner
Web-scale search, big data, and ML leader; CTO at Dropbox and ZoomInfo; Illinois PhD.
Working with us
Three ways in.
We make your conviction stronger.
Join as a subscriber
Send what you're seeing
Hear from the quorum
Invest together
A neutral second opinion from operators — not a verdict over your shoulder.
People who speak your language.
One technical conversation
The quorum engages
A check and a network
Experts & buyers, ongoing
Diligence by people who've built your stack — and introductions to the people who'd buy it.
Your judgment, on the cap table.
Join the network
Contribute where you're deep
Earn co-invest rights
Advance — access, carry, ownership
A structured path from operating expertise into evaluating and investing.
“We're building the venture network we wish we had when we were building.”