The Agentic AI Startup Studio

We build AI-native companies for enterprise-grade challenges.

We Found, Co-Found and Re-Found AI-native companies, built with enterprises on a shared agentic platform and operated by us from day one.

#01 / What we do

We are owner-operators, obsessed with hard problems and the long, unglamorous work of building durable companies.

We don't write checks from the sidelines. We start companies, take operating seats, and stay in them, building on a shared AI-native platform so every venture compounds the last.

Own it

Skin in the game from day zero.

We hold founder and operator roles at inception, with capital, reputation, and time committed alongside the companies we build.

Solve it

Obsessed with the problem, not the pitch.

We pick problems worth a decade of work, validate them with real customers from the first week, and refuse to build companies the market doesn't pull.

Compound it

Built to last, not to flip.

We architect ventures for durable margins, defensible data, and the kind of operating depth that turns first products into enduring companies.

#02 / Three modes

Where companies come from.

Three distinct origination paths into the studio, each with its own deal shape and founding team.

Mode 01

We Found

We incubate companies alongside an enterprise client and partner — originating the thesis, building the product on our agent platform, and launching with a first customer in place.

Mode 02

We Co-Found

We co-create ventures with proven operators or product and tech founders who bring the problem and domain expertise, shaping the company together from concept.

Mode 03

We Re-Found

We acquire and rebuild under-used technology as a modern, agentic AI company, applying fresh operating focus and a new go-to-market to unlock latent value.

#03 / How it works

What happens after we say yes.

We seed every company with capital, the platform, and an operating pod. In return, the studio holds founding-team equity and stays meaningfully involved through the first priced round.

Step 01

Commit

A thesis is greenlit — from us, an enterprise partner, or an operator — and pressure-tested against real buyers in weeks.

Step 02

Incubate

We assemble a pod, build on the shared platform, and land the first paying customer before naming a CEO.

Step 03

Spin out

The company raises a priced seed, the operator takes the wheel, and we stay on as engaged founding shareholders.

#04 / Why now

Software is cheap to build. Operating judgment isn't.

01

Capital and code are no longer the bottleneck.

Agentic AI has collapsed the cost of building software. What's scarce now is distribution, proprietary data, and operating judgment, and that's our edge.

02

We're operators, not spectators.

We hold leadership roles at inception, so companies get hands-on product, go-to-market, and operating support, not just advice.

03

Every company starts ahead.

A shared platform, reusable technology, and a growing data advantage mean each launch is faster and better-informed than the last.

#05 / The Platform

Three reusable assets, so every company starts ahead.

Not slideware. Production systems we build once, harden in market, and ship into every new company on day one.

Asset 01

Agentic Harness

A production-grade orchestration layer for multi-agent workflows, with evaluation, observability, and human-in-the-loop control built in.

Ships in weeks, not quarters.

Asset 02

Enterprise deployment kit

Identity, audit, data residency, and security controls pre-wired for procurement at Fortune 1000 buyers, so our companies clear review without rebuilding the stack.

SOC 2 and SSO from day one.

Asset 03

oGraph data layer

A shared, privacy-conscious data cooperative — launched as its own company (see below) and available to every studio venture on a give-and-get model, so each launch starts with signal, not a cold start.

Compounds across every venture.

#06 / For Enterprises

Your agentic innovation vehicle.

Enterprises work with Founders United to turn their hardest problems into new companies, shaped with them from the start, operated by us, and built on production-grade agentic AI. A way to pursue innovation outside internal gridlock, with experienced operators carrying the build.

Explore a partnership
01

Shape solutions from the concept stage, instead of evaluating finished products.

02

Engage experienced builders who carry the operating burden.

03

A practical path to deploying agentic AI in the enterprise.

#07 / For Operators

Build the company you've been waiting to build.

We back proven operators and product or tech founders who want the leverage of a studio without the politics. You bring the conviction. We bring capital, customers, and the platform to compress years into months.

Apply as a founder
Equity

Founder-level ownership, vesting from day one. Real upside, not advisory crumbs.

Capital

Backing to focus on the work, with a clear path from incubation to a funded company.

Day one

Design partners lined up, platform infrastructure live, a pod of builders around you.

What we expect

Full-time, in-market, and willing to take the CEO seat if the company finds its shape.

Companies.

#08 / Portfolio

iCustomer

In Market

Decision intelligence for marketing and go-to-market teams. Turns fragmented customer data into live, decision-ready audiences, with AI agents that streamline the workflow.

Decision Intelligence

oGraph

In Market

A privacy-first consumer data cooperative built for agents. Portable, permissioned identity and signal that compounds across the ecosystem instead of locking into one platform.

Consumer Data Cooperative

Double Black

In Development

A vertical AI marketplace connecting ski brands, retailers, and partners. Solves demand forecasting, inventory optimization, and audience intelligence for a category still run on spreadsheets.

Vertical AI · Commerce

Agency Agentic

In Development

Service-as-software for paid media. A bespoke agency wrapped around an agentic growth loop that learns from every campaign and compounds performance over time.

Service-as-Software · Growth

More ventures in development across commerce and enterprise.

More soon.
#09 / Team

Four founding partners.

Collectively, the team has founded companies, served on technology-company boards, and built and operated businesses across enterprise and commerce.

Bob Geiman
Partner

Bob Geiman

4x founder · venture investor

Predictably Human, Confirm.io · Polaris, Summit, Warburg Pincus heritage · Dartmouth, HBS

Abhi Yadav
Partner

Abhi Yadav

3x founder · author · deep-tech angel

Zylotech, InsuraMatch, Reltio, GE · MIT

Jonathan Seelig
Partner

Jonathan Seelig

3x founder · venture investor

Co-founder, Akamai (NASDAQ: AKAM) & Ridge · Zipcar board · Stanford, MIT

Laurence Constable
Partner

Laurence Constable

2x fund builder · attorney

McCarter & English, Holland & Knight · counsel to founders, funds & family offices

Track record

$10B+ built. 18 founded. 39 exits.

4.5x

Realized return

18

Founded

39

IPOs & acquisitions

85+

Boards & C-suite seats

Select exits
uber.comta-associates.comfiserv.comintel.comjackhenry.comaol.cominterpublic.comlibertymedia.comblackberry.comakamai.comverizon.comgoogle.comadobe.combroadcom.comnokia.commeta.comriverbed.comphreesia.com
#10 / Community

An invite-only room where the right people meet.

The AI Innovators Community (AIC) is a curated network that helps our founders meet enterprise clients, cofounders, partners, advisors, and investors, in person, across San Francisco, New York, and Boston.

Who

Enterprise clients, cofounders, operating partners, advisors, and investors working on agentic AI.

Where

Invite-only gatherings across San Francisco, New York, and Boston.

Why

Warm introductions to the customer, cofounder, or check a company actually needs next.

#11 / For Investors

Back the studio, and every company that comes out of it.

We partner with a small group of institutional and strategic investors who want exposure to AI-native company creation at the studio level, with pro-rata into every venture as it spins out.

Thesis

Agentic AI, enterprise-grade problems, operator-led.

Cadence

Two to four new companies a year, launched with a first customer.

Access

Studio LP position with pro-rata rights into every spin-out.

Talk to us