Erik Gelcich
I take seriously the things people call dumb. That is where the best ideas hide.
I take products from 0 to 1 on my own: I research the behaviour, design it, build it, ship it, and iterate on the data.
- Now living in
- Netherlands
- Languages
- English, Italian, French
- Open to
- Part-time in NL (until Feb 27)
A social feed for the airport gate: everybody is stuck, nobody knows each other, and everyone has a reason to be there.
A party game that lives entirely inside Instagram DMs. No app, no game screen - the website is only the lobby and the scoreboard.
Write a message, get back a random AI distortion of it. Not knowing what comes out is the whole product.
A personal agent that runs my day: unstructured input in, structured tasks and calendar blocks out.
Social media for right here, right now. 1,000+ users and 800+ posts in the first month, live at Aalto University.
Live audience voting for pitching events, and a growth channel for Mimmo.
Audience-driven Q&A for panel talks. 50 questions became a 2-question flow.
Song requests DJs actually want, collected before the gig, with no signup.
All-in-one practice companion for conservatory music students. Tested by 80+ musicians.
Gate
2026A social feed for the airport gate: everybody is stuck, nobody knows each other, and everyone has a reason to be there.
A web app for the wait at the gate: post, and meet the people waiting on the same floor. An AI asks follow-up questions to turn a thin post into a real one.
The gate is the last point where airports and airlines have a captive audience, and it goes unused. Passengers stare at departure screens, watch the boarding door, or refresh flight status.
External research on boarding (TNMT, tnmt.com/boarding-is-aviations-overlooked-innovation-frontier) breaks the wait down: 36% of passengers wait at the gate and 17% are on their phones, against 27% who dine or shop and 20% who explore. Around half the terminal is doing nothing in particular.
Three things are true of an airport at once: it is dense with strangers, anonymity costs nothing because nobody stays, and if you are flying there is usually a reason. The reason is the product. Airports are full of stories nobody tells, sitting next to people who would listen.
A feed of short posts, plus a section for meeting people in the terminal at the same time. Onboarding is a QR code in the airport: no install, no account beyond a username and where you are flying. A blank composer gets blank posts, so an AI asks follow-up questions and draws out what a poster would leave out.
1,000 people on the first day, and 38 posts. Big arrival, small conversion: reading costs nothing, posting still asks for something. The by-product was as useful as the feed. Destinations plus posts give an airport a live read on what passengers feel and what is going wrong in the terminal, which a survey sent after the flight cannot.
The only way in was a QR poster in the terminal, which put distribution inside somebody else's building. Most of the posters came down and usage collapsed behind them. Onboarding that depends on physical space you do not control is the same trap as building on somebody else's API, except a poster can be taken down by a person who never heard of the product.
An AI-native CRM for HMD's sales team: less data entry, more selling.
TrailRadar: AR spectating for ultra-trail races, built on EU space technology.
Real-time asset awareness for power-grid field crews. The VEO challenge.
A social platform you can only join while standing in the place it belongs to.
Harvia's IoT sauna sessions turned into a cooperative mystery game that builds a routine.
Prompt
2026An AI-native CRM for HMD's sales team: less data entry, more selling.
With two teammates, we started from the salesperson's real problem: a CRM demands data entry and gives little back. We designed an agentic CRM that reads your screen, your past actions and the context, then recommends the next action. Less a database, more an analyst on the team. Won the HMD challenge.
Aalto's intensive venture-building program for the most driven founders-to-be.
Led ski groups of up to 13 people in English, Italian, and French - from kids to seniors.
Private chauffeur for a night club, working with renowned international DJs, like Bob Sinclar, Deborah De Luca, Hugel and many more.
Independent music production: composition, sound design and mixing.
Performed at bars, clubs, and sport events across Italy and Finland.
Front-of-house and bar work in restaurants, bars and a night club.
Founder Sprint
2026Aalto's intensive venture-building program for the most driven founders-to-be.
A 15-week programme of deep-dive modules, coaching and real venture building. This is where Mimmo was born. I worked in teams of 13 and 4, coordinated planning and execution, ran 20+ cold-outreach interviews to validate demand, and pitched our progress to venture capitalists, startup mentors and industry experts.
Need one person to design it and ship it?
Not the right fit for your team?
Share it with someone in your network who might be interested.