The job posting reads like a dare. Pick a startup idea. Build a working version. Record what you made. Send it in. If Persist likes the result, the reward may not be a gift card or a polite recruiting call. It may be a salary, an equity stake and the responsibility of running the company. Persist calls the format Startupathon. The name suggests a weekend of cold pizza and hurried code; the proposition is much more consequential. It is an audition for ownership.
That inversion explains the Los Angeles company better than its broad portfolio ever could. Persist describes itself as a venture studio and capital platform. Its companies range from face-recognition search and an AI companion to remote recruiting, creator software and household robotics. At first glance, the collection looks like an overstuffed ideas notebook. Look again and the common product is visible: a repeatable system for discovering people who can turn an assignment into a business.
A company that manufactures companies
Traditional venture capital waits for a team to form, a product to appear and a pitch to arrive. A studio begins earlier. It develops theses, supplies operators and owns a meaningful share of what gets built. Persist adds a talent competition at the front. Candidates choose from prepared concepts - or sometimes propose their own - and submit an MVP with a demonstration. Reviewers offer feedback, and the cycle continues until somebody looks capable of leading the project.
The Persist loop
The mechanism solves two stubborn problems at once. The first is access. Promising builders often live outside the few schools, employers and cities that make investor introductions easy. The second is signal. Entrepreneurship is difficult to infer from a résumé. By asking everyone to solve the same concrete problem, Persist gets evidence: speed, taste, judgment, technical range and the ability to explain a decision. A prototype is imperfect evidence, but it contains more information than enthusiasm alone.
“The first to deliver gets funded, gets equity, and gets crowned CEO.”Persist's description of Startupathon
The studio is the hidden product
Once a leader is selected, Persist's other identity comes forward. The company says it provides designers, developers, recruiters, marketers, advisors, internal tools and access to capital. Its careers pages recruit venture builders alongside finance leaders, growth specialists and Web3 operators. Its partner program advertises agency work including development, recruiting, search optimization and press, with referral fees and the possibility of joint ventures.
This makes the business model a hybrid. Shared services can earn fees. Referrals can earn commissions. New companies can give Persist equity. A portfolio winner can become far more valuable than years of service revenue, while service work helps keep operators busy between launches. The arrangement resembles a small conglomerate designed around reusable startup labor: one recruiting function, one design bench and one distribution playbook spread across many experiments.
Culture, compressed into a deadline
Venture studios often describe culture with familiar nouns: ambition, curiosity, grit. Persist translates those nouns into a work cadence. Challenge instructions emphasize a real assignment, a visible submission and feedback soon after. Candidates are encouraged to document progress, explain their choices and return with a sharper version. The habits are closer to a newsroom or a product lab than a classroom. The work is public enough to create pressure and concrete enough to make disagreement useful.
The company repeatedly invokes ownership. That word carries two meanings here. There is literal equity, which Startupathon advertises alongside salary and initial company funding. There is also an operating expectation: do not wait for a perfect specification, find the next constraint and move the product forward. Persist's line “ownership > salary” neatly joins the financial incentive to the desired behavior. The company wants the person making day-to-day trade-offs to think like the value belongs to them because, in part, it does.
A global team makes the system more demanding. Persist's public company profile has placed its workforce in the 201-to-500 range, while outside company data supplied for this profile estimates roughly 170 people. Many team members are based in India, even as headquarters remain in Los Angeles. Shared documentation, short demo videos and repeatable review criteria are not decorative process in that setting. They are how a distributed studio keeps dozens of simultaneous bets legible.
The culture also has an ideological streak. Persist talks about altruism, positive impact and helping humanity flourish, while its historic portfolio grouped projects under headings such as “Web3 for Good,” “Redefining Institutions” and “Fighting Bad Habits.” Some ventures fit that language cleanly; others are simply commercial software. The tension is revealing. Persist is trying to run a factory without sounding industrial - to standardize company creation while leaving room for founders to care deeply about what comes off the line.
Figures above are reported by Persist on its website. The company does not publish the portfolio-level methodology behind them.
What customers actually buy
Persist serves several groups, although they do not all pay in the same way. Aspiring founders get a structured route into entrepreneurship. Existing founders can use product, hiring and growth support. Service clients can buy agency work. Investors gain exposure to a collection of early companies. The ultimate customers sit one level down, inside the portfolio: a person checking where a photograph appears online, a hotel considering an AI concierge, a business hiring remote talent or a household curious about a more capable cleaning robot.
Three ventures on Persist's current homepage make the range tangible. Face Search AI lets a user upload an image and look for appearances of that face online; Persist says the product has served more than one million users. Swissmote recruits remote talent, primarily from India and other emerging markets. Open Droids is developing home robots for cleaning, hospitality and elder care. Founder Jackson Jesionowski - who also uses the name Jack Jay - has stepped back from day-to-day leadership of Persist to run Open Droids full time.
Illustrative comparison of hands-on company formation - not a financial measurement. Persist combines capital, training, recruiting and operations inside one system.
A different kind of moat
Persist competes with several categories at once: accelerators such as Y Combinator and Techstars, founder-matching programs such as Antler, venture studios such as Atomic and High Alpha, plus recruiters and hackathons. Its distinction is the sequence. An accelerator generally selects an existing founding team. A recruiter fills a defined role. A hackathon rewards a project. Persist attempts to use the project to select the founder and then wraps a company around both.
The defensible asset, if the system works, is not any single challenge. It is the accumulated operating data: which briefs attract strong builders, which behaviors predict leadership, how quickly shared teams can ship and which distribution tactics transfer between products. Each experiment can make the next one cheaper. The model also creates an unusually direct recruiting message. “Come interview” is ordinary. “Come build the company you might own” is not.
There are trade-offs. A fast prototype rewards speed and fluency with modern tools; it may miss patient researchers, careful enterprise sellers or leaders whose advantage appears only over time. A studio-originated concept may produce a capable operator without the deep personal obsession that carries founders through a decade. And a portfolio spread across AI, consumer social products, crypto and robotics can strain even a large shared team. Reuse is powerful until every company needs something different.
The wager behind the spectacle
Persist's own story began around 2017 with Jesionowski's interest in founder fellowships. The company says he had applied to such programs and, after not being selected, began building projects and eventually a structure for supporting other ambitious people. That origin matters. Startupathon is not merely a loud hiring funnel; it is an argument about missed talent. The studio believes founder potential is distributed more widely than access to founder capital.
That is why India features prominently in its recruiting story. Persist has described the country as rich in technical talent and comparatively underserved by venture networks. The pitch is economic, but also social: connect builders who can execute with advisors and investors they might not otherwise meet. Public challenge pages have ranged from adaptive education and collaborative music to automated app creation, voice hospitality and self-awareness software. The briefs are sometimes strange. Strange is useful; it makes rote answers harder.
A pitch deck records conviction. A build challenge records behavior. Persist is betting the second is a better place to begin.
The company reports 30 launches, a 400-person advisor network and $117 million in net asset value. Those are company figures, not a full window into returns. Earlier public descriptions also claimed two exits. What can be observed more directly is the volume of concepts, completed challenges and named operators moving through the system. Persist is not hiding the messy front end of company creation. It has turned that mess into the main event.
For a would-be founder, the offer is practical: choose a real problem, demonstrate what you can do and potentially inherit more infrastructure than a solo builder could assemble. For a company seeking help, Persist offers a menu of operators and services. For the wider market, the experiment asks a useful question. If venture capital keeps drawing founders from the same social graph, what happens when the application is not a pedigree or a pitch - but a thing that works?