Dreamups makes the first startup less of a lonely experiment

Linotake began with two things on its menu: restaurant reservations and food delivery. By August 2020, its founders had removed delivery. In an interview during Dreamups’ Upcelerator program, Lilia Oglinda and Dumitru Slonovschi described concentrating on reservations and revising features around customers’ requests. An entrepreneurial education had produced a smaller idea. That is an outcome worth paying attention to.

Dreamups is a startup education and acceleration organization in Chișinău, Moldova. Its audience stretches from high school pupils to professionals building their first technology business, then to teams with a working product. The connecting problem is practical: knowing how to build software does not automatically tell you whose problem to solve, which feature to build first, or how to charge.

A campus built around the missing middle

Co-founder Mihai Stipanov dates Dreamups’ founding to 2015. In a 2019 interview, he described an early Moldovan technology scene with important pieces missing. Infrastructure and talent had improved, he said, while product-building knowledge and attitudes toward failure still needed work. His own unsuccessful ventures supplied an unusually candid qualification.

“I failed a lot, and this is my unfair advantage.”

Mihai Stipanov, co-founder · 2019 interview

By 2016, Dreamups was introducing Founder Institute training to Moldova. Contemporary reporting explained that two Dreamups co-founders had completed the institute’s San Francisco program the previous year. The curriculum combined weekly sessions with business-building assignments. The import was a method people could practice, with experienced founders close enough to question.

That distinction helps explain Dreamups’ position today. It sits between education and early investment, offering several entry points into the same community. Moldova also has alternatives such as XY Partners and Yep!Moldova. Dreamups’ particular proposition is the connection between school-level exposure, idea-stage work, product training and traction support. A founder can seek different help as the questions change.

The first customer gets a vote

Upcelerator advertises a free, equity-free, three-month online program for Moldovan professionals and aspiring founders, including people living abroad. Its opening customer-development module asks participants to interview potential customers and identify early adopters. Product decisions follow those conversations. Further modules address competition, revenue, sales, finance and fundraising. The sequence makes the customer an early participant in the design process.

Linotake supplies a modest, concrete example. Its founders said they narrowed the project and concentrated on features customers wanted for reservations. Their account describes a pivot rather than a business collapse. The original scope lost its place in the plan. Readers can copy the discipline: identify a specific problem, ask the people experiencing it, and let their answers change the build list.

The bargain also includes effort. Upcelerator says participants who fail to engage may be removed. A remote program makes attendance geographically easier; it still requires time for assignments and decisions. Someone seeking a library of reassuring videos will encounter work that asks for evidence.

Six months, and a price beyond money

UpNext addresses a later stage. Its six-month traction accelerator combines individual mentoring, operational support, coworking and investor preparation. The 2026 recruitment notice sought teams with a working minimum viable product, early customer validation, technical leadership and full commitment. Those conditions matter: a vague idea and a team already testing sales need different conversations.

The public price deserves careful reading. UpNext’s website lists €10,000 for 1-5% equity plus performance-based investment. Its 2026 LinkedIn recruitment copy specifies €10,000 for 2%. A founder should establish the agreement for their edition before joining. Additional capital through Mozaic is an opportunity requiring further decisions. A headline funding ceiling is a poor substitute for understanding the share of the company being exchanged.

The selection process became more pointed in 2026. Ten teams entered the first stage from 45 applications, with five planned to advance after a review roughly two and a half months into the program. Execution and product-market fit were named as criteria. The design makes progress consequential while giving more teams an initial period of support.

Teach the people who will build the product

Dreamups also works upstream of the founder’s pitch. Startup School uses teachers to guide pupils through online lessons and team assignments. Its public materials offer Romanian, Russian and sign-language lessons. That arrangement puts guidance inside schools while allowing a shared curriculum to travel beyond the capital. The teacher becomes a local point of contact for unfamiliar business concepts.

Pivotal Product School serves career changers, professionals and companies developing product talent. Its eight-week remote format combines modules with practical case work. In May 2026, it reported 30 graduates from 33 admitted participants; 109 people had applied. Teams worked on case studies involving the government’s EVO service, Selftalk and Brizy. Twenty-six full scholarships helped cover participation.

These programs reveal the economic mix behind Dreamups: partner-supported education, tuition where seats are self-funded, and investment-for-equity acceleration. Partners include UMAEF, Innovate Moldova and the Mozaic investor community; Orange has supported earlier education initiatives. The learner, the sponsor and the investor occupy different positions. Their money should be understood accordingly.

Blueprint interviews and Startup Grind events keep founder experience circulating outside formal cohorts. For a newcomer, they offer a way to hear what operating a company involves before applying. For another community building founder support, Dreamups suggests a useful principle: give people an appropriate next task, then put someone nearby who can challenge the answer. Linotake’s shortened menu makes the point neatly. Sometimes the next task is deciding what to stop building.