Crew Live · iOS & Android
A shared home that runs itself. Split what the flat spends, divide what it has to do, and stop having the same argument about money every month.
Product · Founder · Milan
I keep building products that take the work off people.
From the first interview to the store listing.
A shared home that runs itself. Split what the flat spends, divide what it has to do, and stop having the same argument about money every month.
Plan the week, and never write a shopping list again. Farro turns the meals you chose into a list sorted by aisle and scaled to the number of people at the table.
It sits in health, and the interesting part is exactly the part I can't show yet. Ask me in a few months.
I can carry a product from a blank page to the store on my own. The research that decides whether it deserves to exist, the calls about what it actually is, the design, the development, the release, and everything that comes back afterwards. I've done it inside a team and I've done it alone, and the good version of this job needs both.
What follows is roughly the order it happens in. The one thing that never moves: I never get to hand a decision to somebody else and stop thinking about it. Every call stays mine to defend, which is the slower way to work and the reason I trust what comes out.
Before anything gets designed I go and find out whether the need is real: interviews, mining the reviews of everything adjacent, market and industry data, and a demand check that doesn't lean on my own enthusiasm. I've built my own internal tooling for that research, because the slow part was never the thinking. Most ideas don't survive this. That's what it's for.
Crew runs my flat, so I catch its bad days myself. That's a head start, not evidence: being your own user is the most biased input in the room, and it gets checked like any other. So I run the interviews, mine the reviews and watch what people do rather than what they say they want — on every product, including the ones I open every morning. Validation is what turns a hunch into a decision.
Before I design an interaction I go and find who solved it well already: the platform guideline, the research, the product that got it right. Apple's Human Interface Guidelines, Material, Nielsen Norman. They're free, and most of what looks like taste is somebody having read them. Then I hold my version to that bar instead of to my own opinion.
A swipe that doesn't follow your finger is not a smaller version of a swipe. It's a broken one. Direct manipulation is a documented expectation rather than a preference: the platform guidelines ask a gesture to track the finger and carry its momentum, and people feel it in the first second when it doesn't. That last twenty percent is the whole difference between an app that feels expensive and one that doesn't.
A toggle in settings is often a decision nobody wanted to make. Every added choice costs the person a moment of thinking — that's Hick's law, and it's why the strongest products ship an opinion instead of a preference. When a screen refuses to come together, the problem is rarely the layout. It's that nobody has said out loud what the thing is for.
Nobody remembers the feature you left out. Everybody feels the one you rushed. Working on my own, scope is the only real lever I have, so I pull it early and without ceremony. Half a finished product beats a whole unfinished one, and it costs less to get there.
It's in my workflow every day, and how exactly is the part I keep. What I'll say is that it moved me from writing specs to shipping the thing the spec described. Open for the longer answer
I don't just use what the industry ships. I build my own tooling around it — internal systems, prompts and small agents that hold my standards so I don't have to re-explain them every week. Keeping up is part of the job now, and most of the value is in the plumbing nobody demos.
It runs across the whole thing, not one stage of it. Desk research and competitor teardowns. Turning messy interview notes into something I can argue with. Drafting and rewriting copy until it says one thing. Producing the screens. Getting the build moving, then testing it and hunting the edge cases I'd otherwise find three weeks late.
The honest part is that it also raised the floor for everyone, which means the floor is no longer where you compete. When anybody can produce a passable screen in an afternoon, the difference moves to the things it can't hold: knowing which problem is worth the quarter, having sat with enough users to smell a wrong answer, and being willing to throw away work that came out fine but shouldn't ship.
What it doesn't do is decide. It made the cost of trying an idea almost nothing, which means the scarce thing is no longer execution, it's knowing what deserves to exist. That's the part I'm actually paid for, and the part I'd never delegate.
I left school at seventeen to chase acting. Four years of theatre and voice work, and a version of myself I had to invent from nothing to pull it off. What it gave me is the thing I still use most: how to get inside somebody else's head and stay there. To read what a person needs from the way they behave rather than from what they say they want.
Then 2021 arrived and the stage stopped. I wanted to make things that outlast a run, and I had no product degree, no network and no shortcut. So I studied online. FutureLearn first, then Coursera, Udemy, edX. Certificate after certificate: business, project management, product, entrepreneurship. It wasn't a title that got me my first opportunity. It was knowing the job before anybody had hired me to do it, which is why I still think online learning is the most underrated thing in this industry. I went back for the conventional route too, later: the diploma, then university.
Books did the rest, and still do. Product Management in Practice, Continuous Discovery Habits, Inspired, The Mom Test, Don't Make Me Think, Zero to One, The Outsiders. Plenty from outside product too, because the job borrows from everywhere. Reading somebody else's experience is the cheapest way I know to buy time.
My first product role was at Contents, where I owned the B2C side. An AI company in the months the category went public: nothing was settled, most decisions got made twice, and product went from a job title to the thing I'm actually interested in. Then a spell at Fiscozen, one of Italy's fastest-growing fintechs, where I arrived improvising and left with a method — how a scaleup prioritises, ships, and says no.
Then came the part that changed everything: I started building my own. Crew was the first. Nobody hands you a brief when it's yours — you find the problem, you set the price, you negotiate with developers and stores and people who will never read your changelog, and every one of those conversations is a skill no job description would have given me. It taught me more than anything else I've done, and it's still teaching.
So that's the plan now, deliberately: a portfolio of apps and software that actually work for the people using them. Not a shelf of side projects. And alongside it, the work I take on with other teams. I'm picky rather than closed: I want to give my time to the projects I'd still be thinking about on a Sunday.
Meal planner with an automatic, aisle-sorted shopping list. Shipping now.
Consumer app for shared living. Research, product, UX, go-to-market and release.
Product initiatives and key user flows inside a fast-moving fintech scaleup.
Owned the B2C product end to end at an AI company, through the period the category went public.
Internship. My first look at how a large organisation actually decides anything.
If it's the kind of thing I'd still be turning over on a weekend, I'd love to hear about it.