What does AI-native mean?

It means the studio practices its own thesis. Our research, diligence, and company-building run on AI with partner judgment on top, the same architecture we ask of every portfolio company. In practice that shows up in the work: market maps and technical diligence drafted by models and corrected by partners, operating memos written once and maintained continuously, internal tooling built rather than bought. It also describes what we build: companies designed around AI economics from day zero, with margin structure, headcount plan, and product surface assuming model capability rather than retrofitting onto it later.

Do you have an application deadline?

No. We run an open application policy. Applications are read on a rolling basis, so there is no window to miss and no penalty for writing to us between cohorts. Every application is assessed case by case on its own merits rather than scored against a fixed rubric or ranked against the other submissions that happened to arrive the same week. If the fit is right and the timing is not, we say so and keep the conversation open.

What stage do you accept?

Idea stage through seed. Most of what we admit is pre-product or pre-revenue, because that is where a co-founder level of involvement changes the outcome most. If you are further along and the fit is obvious, write to us anyway: applications are open year round and assessed case by case, so stage is a starting point for the conversation rather than a filter that runs before it.

What happens after I apply?

A partner reads every application. Within about two weeks you hear back either way. Promising applications move to a first call, then to a working session where we take a real problem from your business and work it through together, which tells us more than a pitch does. Diligence and terms follow for the companies we decide to build with. We aim to give a clear answer quickly rather than leaving founders in an indefinite process.

Does Centaur Labs invest capital?

Yes. The Labs invests at formation and leads the first check into every company it builds or admits. We are not a fee-taking accelerator and we do not charge for participation. Terms are shared during diligence, before any commitment is asked of you, and they are the same structure for every company in a cohort. We also work alongside founders on the round that follows, including introductions to the funds most likely to lead it.

What do you look for in a team?

Accountability over capability. Model capability is widely available and improving on someone else's roadmap; what stays scarce is a team that owns the outcome, decides quickly, and tells the truth about what is working. Beyond that we look for real proximity to the problem, evidence of shipping under constraint, and a willingness to build in the open with a partner sitting next to the work rather than reviewing it quarterly.

Do you also start companies yourselves?

Yes. Roughly half the portfolio is studio-originated: theses we develop in-house, validate with operators in the field, and then build with recruited founding teams. Those companies get the same capital and the same partner involvement as admitted ones. If you would rather join a company than start one, say so when you write to us and tell us which of the four fronts you want to work on.

Do you accept solo founders?

We prefer them. The Labs is designed around the assumption that you are one person with the leverage of a firm; our diligence tests judgment, not headcount. Teams are welcome too, provided they are small enough to move like one.

What if I do not have a co-founder yet?

Apply anyway. Solo founders are welcome, and a missing co-founder is a solvable problem rather than a disqualifying one. We work with founders on the search, and where the fit is right we match applicants to studio-originated theses that already have a team forming around them. What we do not do is manufacture a partnership for the sake of the cap table.

Where is the program based?

Washington, DC, with participation designed for founders anywhere. The working rhythm is remote by default with periodic time together in DC around the moments that benefit from a room: formation, the first customer push, and the raise. Founders outside the United States have run the program without relocating, and we help with the practicalities where we can.

How much time does the program take?

The cohort runs across a defined arc rather than a fixed number of contact hours, and we expect founders to be working on the company full time. In practice that means a standing weekly session with a partner, working blocks around whatever the company is actually facing that month, and access between sessions rather than office hours you have to book. We would rather be useful on your schedule than fill a curriculum.

What happens at the end of the cohort?

Nothing ends. The formal arc closes with a partner-led introduction process into the funds and strategics most relevant to the company, but the studio stays involved as an investor and as a working partner for as long as it is useful. Follow-on capital is decided company by company on the evidence, not promised in advance.

Do you sign NDAs at the application stage?

No, as is standard for investors. Signing NDAs across a rolling application flow would make it impossible to look at adjacent ideas honestly. In exchange, every application is treated as confidential, read only by partners, never circulated to other founders or funds, and never used as raw material for a studio-originated thesis. If there is a specific piece of information you are not ready to share, leave it out of the application and raise it once we are talking.

The Centaur Labs · The first AI-native venture studio

01 · THE THESIS

Intelligence is cheap. Judgment is scarce.

Every six months the models get smarter and the tokens get cheaper. Capabilities that were research problems in 2023 are commodities in 2026. Whatever a company does with raw intelligence alone, a competitor will do next quarter for less.

And yet the markets that run on trust, the ones where a professional signs the work, a patient opens up, a reader believes, and a buyer commits, remain barely transformed. Not because the technology is missing, but because these markets do not buy capability. They buy accountability. A professional cannot bill for work they cannot defend. A person will not hand their inner life to a system no one answers for. A reader will not believe a masthead no one signs.

That is the opening. The durable AI companies of the next decade will not be the ones that remove humans from the loop. They will be the ones that redesign the loop: machine scale underneath, human judgment on top, and a clear account of who decided what, and why. Kasparov called this pairing a centaur. On the chessboard, it beat the machine and the grandmaster alike.

We believe the same holds in the economy. Centaur Labs exists to build centaur companies. And the purest centaur of this era is the solo founder: one judgment, machine everything else.

02 · THE SOLO FOUNDER ERA

One founder is now a firm.

For fifty years, the founding team was a technology constraint. You needed an engineer to build, a designer to ship, a marketer to reach anyone, an operator to keep the wheels on; even the myth of the garage had two people in it. Venture capital priced the constraint in and then enforced it: the solo founder was a red flag, a diligence checkbox, a discount on the term sheet.

The constraint is gone. A single founder with today's models writes the code, drafts the contracts, produces the content, answers the customers, and closes the books. Agents work through the night. The marginal cost of a colleague has collapsed, and with it the old arithmetic of the founding team. The companies of this decade will be built by the smallest teams in the history of the industry, and some of the best of them will be built by one person.

What one person cannot manufacture is exactly what the machine cannot provide: a second judgment, an honest argument, a network that answers the phone, a name beside yours when the institution across the table asks who else believes. Co-founders used to supply this. Now a studio can, without taking half the company to do it.

That is what Centaur Labs is built for. Our diligence assumes you are alone, our support replaces the org you did not hire, and our first check is priced on conviction, not headcount. You bring the judgment. We bring the firm.

“The question is no longer whether one person can build the company. It is who stands beside them when it counts.”

03 · THE METAPHOR

A figure older than the industry.

The centaur is one of the oldest images our civilization has of a being made of two natures. In Greek myth he is Chiron: teacher of Achilles and Asclepius, healer, strategist, the rare centaur who tempered animal instinct with human reason.

The medieval knight and his horse were trained until they fought as one body, a centaur reborn in armor. And after losing to Deep Blue in 1997, Garry Kasparov gave the word its modern meaning: a centaur is a human paired with a machine, and in serious chess that pairing reliably outperformed either one alone.

That is the wager of the Labs. The model is the horse: powerful, fast, sometimes wild. Judgment is the rider. We back founders who refuse to let go of the reins.

04 · STRATEGY

How we build.

Centaur Labs is a venture studio, not an accelerator. The difference is not cosmetic. An accelerator processes companies; a studio commits to them. We take few positions, take them early, and work them with the intensity of a co-founder rather than the distance of a fund.

AI-native, from the studio down.

Centaur Labs is the first AI-native venture studio: not a fund that invests in AI, but a studio built on it. Our research, diligence, and company-building run on the same centaur method we back. The thesis work behind every focus area, the competitive maps, the regulatory analyses, the first draft of everything: produced at machine speed, signed by senior humans. We build companies the way we tell founders to build products. A studio that only advises on the future is a consultancy. We would rather be the proof.

Proven models, new physics.

We do not hunt for unprecedented business models. We hunt for proven ones whose economics AI has quietly rewritten. Take any established model: the law practice, the therapy clinic, the newsroom, the executive’s staff, the brokerage. Each is a business the market already understands, already buys, and already prices. And each has been held down for a century by the same two ceilings: expert time is expensive, and expert attention does not scale.

AI removes both ceilings at once. The cost of producing the core service collapses, and the audience that can afford it expands by an order of magnitude. What was a boutique becomes a platform. What served the Fortune 500 can serve the mid-market. What served the wealthy patient can serve the whole waiting room. The demand was always there; it was simply priced out.

That is where we invest: AI-native versions of businesses the world has already voted for, rebuilt with a cost structure incumbents cannot match and a reach they never had. Novel categories require educating the market. Proven categories only require winning it.

Two doors in.

Companies enter the Labs one of two ways. Some we originate: theses developed in-house, in markets we know from the inside, for which we recruit founding teams and build alongside them from day zero. Others come to us: founders already in motion who apply to a cohort and want partners with skin in the argument, not just the cap table. Both doors lead to the same room, the same partners, and the same standard.

Built for one.

Most venture infrastructure quietly assumes a team: the accelerator batch, the co-founder matching, the board-seat arithmetic. We inverted the assumption. The Labs is designed so a single founder operates with the surface area of a firm: partner judgment on tap, specialist advisors matched by focus area, design customers drawn from our network, and an AI-native operating stack we install from day one. Where others diligence your headcount, we diligence your judgment. The machine takes care of the hands.

Concentration over coverage.

A cohort at the Labs is smaller than what most programs admit in a single category. This is deliberate. Advice scales badly; attention does not scale at all. Every company in the portfolio gets a named partner, a standing weekly session, and the right to our honest opinion early, while it can still change the outcome. We would rather be decisive in eight companies than diluted across eighty.

Trust is the moat.

Generic consumer AI is a knife fight over distribution. Markets built on trust are different: professionals with licenses and reputations, patients with their inner lives, readers deciding what to believe, buyers and sellers committing real money on a match. Most builders treat that weight as friction. We treat it as the moat. A product that earns a professional’s signature, a clinician-grade standard of care, or a marketplace’s integrity has built something a weekend wrapper cannot copy. Our edge is that we have operated in rooms where trust is the currency, across forty markets and twenty years, and we know what it takes to earn it.

From first customer to first close.

The studio’s job divides into three phases. Sharpen: thesis, market, and regulatory posture, pressure-tested before you spend a dollar defending them. Build: product developed with design customers drawn from our network, so the first version answers to a real buyer rather than an imagined one. Raise: positioning, materials, and warm introductions made when the company is ready, framed by people investors already trust. The phases are a spine, not a syllabus; every company moves through them at its own pace.

05 · FOCUS AREAS

Four fronts. One discipline.

We build where humans cannot be removed from the loop, only amplified: markets where a professional’s name, a person’s mind, a reader’s trust, or a marketplace’s integrity is on the line. Four categories, one common structure: an established model the world already pays for, rebuilt AI-native, with a centaur at the center. Each is a front where one founder with the right machine can now field an entire company.

Focus area 1 of 4: Media and knowledge

Intelligence, not content.

The content economy is drowning in generated text, and the value is migrating accordingly: away from producing words, toward earning trust and turning information into judgment-ready intelligence. The established models, the newsroom, the research desk, the analyst shop, the newsletter, still command attention and subscription revenue. Rebuilt AI-native, a two-person desk now does the work of forty, publishes in multiple languages, and holds a standard because humans set it.

WHY THE CENTAUR WINS HERE

Distribution now punishes slop; platforms and audiences alike are learning to select for accountability, for a named editor who stands behind the output. Pure AI media converges on noise, and pure human media cannot match the cost base. The centaur newsroom takes both sides of the trade: machine speed in production, human judgment on the masthead. The economics of software, the credibility of a byline.

THE SOLO OPENING

One editor with an agent newsroom can run a masthead that used to take a floor.

Focus area 2 of 4: Wellness

Depth of care, at the price of an app.

Demand for care has outrun the supply of clinicians in every market on earth: waitlists measured in months, sessions priced beyond most households, and a generation fluent in talking to software about their lives. The established model, reflective, methodical care for the inner life, is proven beyond argument. What AI changes is who can afford its first layer: an always-available, method-grounded reflective companion at consumer prices, with human care where it belongs.

WHY THE CENTAUR WINS HERE

The inner life is precisely where naive automation fails and where the cost of failure is highest. Regulators have noticed; the rules arriving across the US and Europe all point the same direction: AI may support, humans must remain accountable for care. The centaur design, machine memory and consistency underneath, clinical method and human escalation on top, is not just the safer architecture. It is rapidly becoming the only one that will be legal.

THE SOLO OPENING

One clinically minded founder can now build the reflective first layer of care for a million inner lives.

Focus area 3 of 4: Tools for professionals (B2P) and productivity

Software for people whose name is on the work.

The business-to-professional economy: lawyers, physicians, analysts, architects, accountants, agents, engineers. People who buy tools not for entertainment but for progress, in their craft, their practice, and their careers. For thirty years enterprise software served the organization and left the individual professional with generic tools and a browser full of tabs. AI flips the equation: for the first time, one professional can afford what used to require a firm around them, research staff, drafting support, a back office, at the price of a subscription.

WHY THE CENTAUR WINS HERE

A professional cannot delegate accountability. The license, the reputation, and the liability stay human, which means full automation is a non-starter and a raw chat window is not a product. The winning tools are centaurs by construction: the machine does the volume, the professional does the judgment and signs the work. Products that respect that line earn trust and pricing power. Products that cross it get banned by the bar association.

THE SOLO OPENING

One professional who has felt the pain can build the tool for everyone who shares it.

Focus area 4 of 4: AI-native marketplaces

Markets where the matching is the product.

Every marketplace is a matching problem wrapped in a trust problem. The last generation solved listings and payments and left the hard parts crude: search plus star ratings on one side, expensive human intermediaries, brokers, agencies, expert networks, on the other. AI-native marketplaces underwrite both sides of the trade: they understand what the buyer actually needs, verify what the supplier can actually do, and often generate the first mile of the work itself. The established models, brokerage, staffing, expert networks, agencies, already move trillions.

WHY THE CENTAUR WINS HERE

A pure-AI marketplace collapses into a content farm; a pure-human one cannot scale past its operators. The centaur marketplace splits the work correctly: machines handle discovery, matching, and quality control at scale, while humans supply the judgment that is actually being bought and sold. Take rates once justified by phone calls get re-earned by outcomes, and the trust layer, verified humans, accountable results, becomes the defensible asset.

THE SOLO OPENING

One operator can run both sides of a market that used to need an office of brokers.

NOW ACCEPTING · COHORT 2026

06 · PORTFOLIO

Where the thesis stops being an argument.

The first companies of the Labs cover three of the four fronts, with two more proving the model in adjacent arenas: research and learning. Some are studio-originated, some founder-led; all are built AI-native, on proven models, with the cost curve on their side.

PRODUCTIVITYLIVE

Everyone has an AI for work. Centaur is the first one for your whole life. It connects your career, your health, and what you’re learning into one system, one score, and puts AI agents to work raising it. Small daily gains, compounding into someone unrecognizable a year from now.

gocentaur.ai

WELLNESSEARLY ACCESS

A psychodynamic companion for understanding yourself. insite turns journaling, dreams, and daily patterns into a private map of the inner life: depth where the market settled for mood tracking, built for the regulatory reality of AI in mental health.

helloinsite.ai

RESEARCH & INTELLIGENCELIVE

Expert intelligence at AI speed. Enquire moves investment, consulting, and policy teams from open question to decision-grade understanding, pairing vetted human expertise with structured AI research in a single environment. The centaur method, sold as a product.

enquire.ai

EDUCATION & LEARNINGLIVE

English, absorbed from life itself. Voscreen teaches language through thousands of short video moments, tuning to each learner and building fluency by instinct rather than instruction, at the pace of play.

voscreen.com

MEDIA & KNOWLEDGELIVE

A global news and intelligence network. Atlas360 turns world news into decision-grade intelligence across newsletters, sites, and podcasts, with AI-native editorial operations and named editors on every masthead. Local desks from Washington to Istanbul, built on a fraction of a legacy newsroom’s cost base.

atlas360.news

Building in one of these arenas, even if it is just you? The next card could be yours. → Apply to Cohort 2026

07 · WHAT FOUNDERS GET

Partners, on your side of the table.

For the solo founder, this section is the point: everything a co-founding team would bring, without the cap-table arithmetic.

Senior judgment.

Weekly partner time with people who have built, advised, invested, and testified. No analyst hand-offs, no junior intermediaries, no office-hours theater.

Capital relationships.

Warm, specific introductions to operators, family offices, and institutional investors, made at the right moment and framed by people those investors already trust.

Global reach.

Forty-plus markets of on-the-ground relationships with regulators, customers, and counterparties. When your first enterprise deal is on another continent, so are we.

Regulatory fluency.

Diligence-grade rigor on what your product, your data, and your story can withstand, applied before a regulator, an acquirer, or a lead investor applies it for you.

08 · WHO SHOULD APPLY

We are looking for riders.

APPLY IF

You are a solo founder, or a team small enough to share one table, with the technology in-house. You are building in one of the four focus areas, from idea stage through seed. Your customer stakes something real on your product: a license, a reputation, a mind, a masthead, a livelihood. You want partners who will argue with you, not applaud you. Solo is not a gap in your application here. It is the design.

PROBABLY NOT A FIT IF

You are building a thin layer on someone else’s model with no thesis on defensibility. You want capital without engagement. Markets built on trust strike you as an obstacle rather than a moat.

09 · COHORT 2026

A clear path, on a short clock.

STEP 01

Apply.

A short written application, read by a senior partner. Not a form that disappears into a database.

STEP 02

Partner review.

We respond within one week: yes, no, or a conversation. Silence is not an answer we give.

STEP 03

Diligence.

Two working sessions on product, market, and team. You will learn as much about us as we learn about you.

STEP 04

Kickoff.

You join Cohort 2026 with direct partner sponsorship and a working plan for your first ninety days.

STEP 01
Apply.
A short written application, read by a senior partner. Not a form that disappears into a database.
STEP 02
Partner review.
We respond within one week: yes, no, or a conversation. Silence is not an answer we give.
STEP 03
Diligence.
Two working sessions on product, market, and team. You will learn as much about us as we learn about you.
STEP 04
Kickoff.
You join Cohort 2026 with direct partner sponsorship and a working plan for your first ninety days.

10 · TEAM

Operators first.

The Labs is run by people who have founded, financed, and advised at the intersection of technology and institutions. We have raised venture rounds, sold to governments and Fortune 500s, taught the discipline in graduate classrooms, and sat through the diligence we now prepare founders for.

CENK SIDAR

Cenk Sidar

FOUNDER, MANAGING PARTNER AND LEAD INVESTOR

Cenk has spent twenty years where technology, capital, and statecraft meet. He founded and led Enquire AI (formerly GlobalWonks), a venture-backed intelligence platform connecting institutions with vetted expertise worldwide, backed by institutional investors through a $5.5 million Series A. Before that he founded a Washington-based strategic advisory firm serving governments, corporations, and investors across forty-plus markets.

He teaches the graduate seminar 'AI Applications in Political Risk’ at Johns Hopkins SAIS, his alma mater, and is the author of Turkish Dream. Affiliations include the Atlantic Council, the Forbes Technology Council, and the Atlantik-Brücke European Young Leaders Program. As lead investor of the Labs, he writes the first check into every company the studio builds or admits.

ADVISORS

The bench.

Every company in the portfolio draws on a bench of senior advisors: clinicians, editors, operators, marketplace builders, and investors who have run the rooms our companies sell into. Advisors are matched to companies by focus area, and they show up in working sessions, not just on the website. The full roster is announced with Cohort 2026.

BACKERS

Patient capital, close to the work.

The first check into every Labs company is our own. Cenk Sidar leads the studio’s initial investment in each company it builds or admits, before outside capital arrives; a studio that asks founders for conviction should show its own first. Around that anchor sits a small group of investors, operators, and family offices who share the thesis. Backer names are shared with founders during diligence.

11 · FAQ

The first
AI-native
venture studio.

Intelligence is cheap. Judgment is scarce.

One founder is now a firm.

A figure older than the industry.