AI for VC, Angel Networks, Incubators, and Accelerators

Your entire deal lifecycle, on autopilot.

FundOS finds startups that match your thesis and diligences them before your panel meets. Then it keeps tracking every company long after you've written the cheque.


How FundOS works

From first signal to funded portfolio

One continuous loop, four phases. Here is every move FundOS makes with your thesis.

Phase 01 of 04
Phase 01 of 04 — Deal discovery

Deal discovery

Source frontier tech before anyone else does.

We pull candidates from Tracxn, Crunchbase and startup databases, run web search in parallel to catch what those miss, then dedupe and score every company against your thesis before it reaches you.

Phase 02 of 04
Phase 02 of 04 — Automated thesis screening

Automated thesis screening

Six core parameters, zero analyst busywork required.

Every sourced company gets scored on six parameters out of the box, and the six are just a default. Tell us what matters to your fund and we'll screen on that instead.

Phase 03 of 04
Phase 03 of 04 — Deep AI investment analysis

Deep AI investment analysis

Institutional-grade diligence ready in minutes, not weeks.

On your top filtered companies, upload the deck and any supporting docs. Thirteen research agents run in parallel — identity, product, people, technology, competitors, market, financials, customer sentiment, brand signals and more — then converge into one scored report.

Phase 04 of 04
Phase 04 of 04 — Portfolio tracking

Portfolio tracking

Real traction metrics beyond self-reported board decks.

Once the cheque clears, connect the tools your portfolio companies already run on and watch real performance against the targets they promised, not just what shows up at the next board update.


Deal discovery, in detail

Tell us the thesis, we handle the funnel

A single search already spans well over a hundred sub-categories inside one sector. Layer on stage, geography, deal size and a dozen more signals, and a raw pool in the thousands narrows itself to the handful actually worth opening.

  • Where to look

    Sector & sub-sectorStageGeographyDeal size range
  • How proven

    Founded-year rangeFunding roundsLatest round sizeAnnual revenue
  • How strong

    Editor’s ratingTeam scoreTracxn scoreInvestment scoreEmployee count

Valuation intelligence engine

Six independent signals, on every company, in under a minute

The scoring layer underneath every stage, built on evidence instead of narrative guesswork. Every source is real, verifiable and cross-checked before we trust it.

  • 01

    Technological moat

    Could a competitor actually copy this, and how hard would it really be?

    Named, TRL-tied government grants found for two-thirds of companies tested

  • 02

    IP strategy

    Does this company genuinely hold registered patents, and what's their status?

    Nine real patents a brand-name search alone would have missed

  • 03

    Team

    Are the founders real, actually at this company, and credible?

    A founder wrongly linked to the wrong company, caught before any spend

  • 04

    Application space

    What is this technology good for today, and where else could it realistically go?

    Six live industries told apart from early, untested potential

  • 05

    Go-to-market

    Who's actually buying, and is that momentum accelerating right now?

    A breakthrough order, a 3× scale-up and a new round, in a single pass

  • 06

    Revenue traction

    What do the real, filed financials actually say, not a narrative someone typed into a deck?

    Filed revenue pulled straight from source, despite a colliding company name


Deep AI investment analysis

Everything the workflow needs, grouped the way you actually work

Not one generalist bot. Focused capabilities across the same three phases your fund already runs on.

  • Discover

    • One Deal, Fully Read

      Signals pulled, fit scored, one clean file

    • Match the Thesis, Not the Noise

      Only startups that fit thesis and cheque size

    • 10,000 Down to 100

      Noise filtered to the shortlist worth your time

  • Decide

    • One Scorecard, Every Deal

      Same framework, every startup, apples to apples

    • Scout to Exit, One Screen

      Every deal stage lives in one place

    • Do More With a Leaner Team

      Research and screening automated, analyst hours cut

  • Track

    • Know the Day It Happens

      Financial tools connected, performance updates live

    • Watch Traction, Not Just Updates

      Real growth tracked, not just self-reported

    • Know Who’s Really Competing

      Real competitors, side by side, always current

    • Hear What Your Sector Is Saying

      Sector conversations tracked before they hit a deck


Why it wins

Built to be trusted, not just fast

  • Every claim traced to a real source

    A filed financial record, a government grant registry, a patent office filing. Never a confident-sounding guess dressed up as fact. Every signal gets independently cross-checked before we trust it.

  • Minutes, not weeks

    What used to be days of manual research on a single company now comes back in minutes, and a full six-signal valuation read finishes in under sixty seconds.

  • Nothing is ever lost

    Every search, every score, every report stays saved and revisitable. Your fund's entire deal history lives in one place, not scattered across old inboxes and forgotten spreadsheets.

  • Scales with the fund's ambition

    A single search can responsibly cover well over a hundred sub-categories inside one sector, as broad or as narrow as your thesis demands, without adding a single headcount.

  • Sourced candidates, narrowed automatically to

    10–30

  • To fully triage 100 sourced companies

    < 1 hr

  • Full 6-signal valuation read, per company

    < 60s

  • Sectors under continuous coverage

    6

  • Independent sourcing channels per search

    2+

Measured, not estimated. Every figure shown here comes off a real pass over a live sector.


FAQ

What partners ask before they trust an AI with diligence.

Before you book a call, this is what comes up first.

Does this replace our investment team’s judgment? +

No. It takes the grunt work off your plate before judgment happens, not the judgment itself. FundOS finds and diligences the companies and builds the scorecard; your panel still decides what actually gets a cheque. Every output comes as a recommendation with its sources attached, never as a decision we've already made for you.

How is this different from a deal-sourcing CRM like Affinity or DealCloud? +

A CRM tracks relationships you already have: who you've met, what stage a deal sits at. FundOS works further upstream. It finds companies you haven't met yet, checks them against your thesis, and runs the full diligence pass before anyone on your team opens the file. Most funds end up running both, with FundOS filling the top of the funnel a CRM was never built to fill.

How reliable is the AI-generated valuation? +

It's not a narrative guess dressed up as a number. Every read pulls from the same six independently verified signals, including technological moat, IP, team, application space, go-to-market and revenue, all sourced from filings, patent offices and grant registries, never from a pitch deck's own claims. Think of it as a fast, defensible first pass your partner still sanity-checks, not a replacement for one.

Can it monitor companies already in our portfolio, or only source new ones? +

Both. This is real portfolio monitoring, not deal sourcing with monitoring bolted on as an afterthought. Track runs continuously on every company you've already funded: performance updates the moment you connect your financial tools, competitors and sector shifts flagged as they happen, not just whenever the next board update rolls around. It's one pipeline covering the whole lifecycle.

We’re moving off a portfolio monitoring tool that’s being retired or outgrown. Is switching painful? +

Not the way people expect. We've talked to funds in exactly that spot, and it's mostly a data-import conversation, not a rebuild of how your team works. Track picks up where the old tool left off: connect your financial tools once, and performance, competitor and sector signals keep updating on their own from there.

What if our thesis covers sectors you haven’t scanned yet? +

A single search already covers well over a hundred sub-categories inside one sector, and we run continuous coverage across six sectors today. If your thesis sits outside that, just tell us on the call. Most expansions turn out to be a config change, not a rebuild.

Does our deal and portfolio data stay confidential? +

Yes, and we know how sensitive this is. Nothing about your fund's pipeline, scores or portfolio data is ever shared across funds or used to train anything for anyone else. Access stays scoped to your own team, and we'll walk your compliance side through retention and access controls before you sign anything.

Stop sourcing deals by hand.

In 20 minutes we'll run FundOS against your thesis and show you the companies it would surface, the signals behind each one and the diligence it would hand your panel. Bring a company you're already looking at if you have one.

No setup needed · Every output comes with its sources