Valas
ENSLBGDE
WhatsApp →
Menu
A senior product team that runs its own products · Slovenia, EU

A working prototype and real traffic.
Not a slide deck.

Seven seniors who ship and run their own products - and answer the question nobody else will: should this be built at all? We answer with working code in front of real users. Then we build only what survives, launch it, and stay until it grows.

Write us on WhatsApp → Where does it hurt? ↓
Start within a day · A demo you can click every week · Stop any time
Where it usually hurts

Six situations we get called into. Pick yours.

Before the first build
Three MVP quotes, €15k to €90k - and nobody asked whether to build it.
A go / change / stop answer in weeks, with a prototype real users have touched.
Should you build it? →
After Lovable, Bolt, Cursor
It works. You have users. You're not sure it's safe to take their money.
Security, payments, tests, EU data - the last 20% that isn't more prompting.
Vibe-coded → production →
After a contractor
Deadlines slipped, then the replies slowed, then silence.
Control of accounts and source first. An honest as-is audit. Then we ship.
Software project rescue →
A live app, a real budget
Installs are cheap. Everything between install and payment is invisible.
Attribution that adds up, a paywall people reach, acquisition that pays back.
Installs ≠ users →
Digital health in the EU
Your third pilot is going well. None of them has turned into a contract.
Pilots with KPIs, trigger events and a budget owner - before day one.
Healthtech pilots →
Inside a company
Twelve pilots, zero deployments, one more deck for the board.
A small fast team that puts a working thing in front of customers and reports weekly.
Corporate spin-offs →
How we work

Test it with code. Build only what survives. Then turn installs into users.

01 - Find out

Name the riskiest assumption. Prototype it in working code - not a mockup. Put it in front of real people and real paid traffic. End with a decision you can defend: go, change, or stop - plus a cut list of what not to build.

02 - Build, same people

The seniors who tested the idea build the product. No handover, no account manager, no "a new person has joined the call". Tests, CI/CD and a clean git history from day one - the product passes a tech due diligence, not just a demo.

03 - Launch & grow

Installs are not users. Growth, store presence, paid acquisition and the paywall are in the room from the first week, so the thing we build is the thing people pay for.

Who you'll work with

You get the seniors themselves - the same people from the first call to launch.

No account managers. No juniors sold as seniors. Nobody changes unless someone fails you.

Andrey Sobolev
Andrey Sobolev
Founder · Product & Strategy
Founder and product lead in regulated domains - telemedicine, B2B, exits. Took over and relaunched a failing consumer app.
Dmitry Popov
Dmitry Popov
Lead Engineer · Architecture
15+ years of architecture in telemedicine, fintech, agritech and consumer products. The engineer who stays calm when production is on fire.
Elizaveta Kook
Elizaveta Kook
Product, Research & Analytics
15+ products, mostly regulated healthtech at tens of thousands of users. Customer research, CJM and analytics that surface what people won't say.
Alyona Semchuk
Alyona Semchuk
Product · Growth, Monetization & GTM
Product manager with 9+ years in growth. Scaled B2C and B2B platforms to tens of millions of users - Windy.app, Outmap, Playorium.
Mykhailo Bochkarov
Mykhailo Bochkarov
ASA / ASO · Mobile Acquisition
Apple Search Ads and ASO for Amazon Shopping, Robokiller, Lotto24, TapeACall and TrapCall. 443K incremental installs, CPA down up to 73%.
Olga Goryacheva
Olga Goryacheva
Content, Product Copy & Messaging
Former editor-in-chief. Writes English product copy as a business tool - for CAC, LTV and product–market fit, not for applause.
Andrey Polugudin
Andrey Polugudin
Senior Mobile Engineer
Swift, Kotlin, React Native. Shipped EatBeat, styledna.ai and mmapp.org from prototype to store - and shut one down when the evidence said so.
Karine Petrosyan
Karine Petrosyan
EdTech Expert · Methodology, Market & Go-to-Market
25+ years as an educator with neurodivergent children - dyslexic herself. Pedagogy and learning methodology at NeuroLearn AI.
Skin in the game

We run our own products. The advice you get is the advice we take.

Every method on this page is one we use on our own money and our own users - consumer apps, AI tooling, a platform that has been running since 2012. When we say "don't build that", it's because we've paid for that mistake ourselves.

Still runningProducts we own and operate today, with live users and live paid acquisition.
ExitedA telemedicine platform taken from idea through a hospital pilot to a strategic buyer.
RescuedA consumer app taken over after another team failed, rethought and relaunched in a new market.
Shut downPrototypes that reached the store and were closed when the signals didn't hold. Weeks lost, not a year.
The fair question

"Why not just Claude for €200 a month?"

Use it. We do - most of what we ship is built with Claude Code. The model gets you from idea to something faster than ever. It hasn't shortened the distance from idea to a good decision.

You're not paying us for code. You're paying for senior judgment, security and accountability for the result - and someone who will tell you what to cut before you pay for it. None of that is a model output yet.

Evidence, not promises

Numbers from products we built, grew or run.

Hospital pilot → exit
Telemedicine B2B2C platform, in production since 2012. Pilot with a major medical centre, exit to a strategic buyer, second product line on the same backend.
443K incremental installs
Large consumer app: +2.23 pp install conversion from systematic creative and store testing; CPA down up to 73% across channels.
Paywall reach 21% → 69%
Onboarding and paywall rework on a consumer app; annual-trial conversion ×1.5–1.7 without new traffic.
+66.2% LTV
GTM for privacy features and a new plan on a top US utility app; a Spanish-language store listing added +5.3 pp install conversion.
Web-to-app: CPI −23.6%
+31% click-to-install for AI apps; a modular creative-testing system took CPA down to −62% and ROAS up 11% across a portfolio.
Days, not quarters
An AI-tooling product of ours went from first server to public release in days - MCP, OAuth 2.1, EU hosting, ~1,000 tests.
All cases →
What we are not

We don't sell hours, and we don't stretch projects.

We don't sell time. Payments are tied to milestones you can click, and you can stop at any of them.

We don't lock you in. Source, accounts, stores and domains are yours from the first commit; when you hire in-house, we help you do it and hand over cleanly.

We don't take equity instead of payment, and we don't take on ideas we believe shouldn't be built. We'll say so on the first call.

Describe the situation in a message. We'll answer with what we'd cut first.

We’re built for fast communication: no forms, no briefs, no call to book. Just write - a senior on the team replies directly.

Write us on WhatsApp → or Telegram
Start within a day · A demo you can click every week · Stop any time · Invoice-based, EU VAT
For more formal inquiries: connect@valas.team