One AI Call. A $20,000 Job.
Most businesses lose their biggest opportunities after closing time. The smartest ones don't.
Turn after-hours calls into booked jobs with an AI assistant that answers instantly, captures every lead, and keeps your business making money around the clock.
Sometimes the most honest way to start a story is with the contradiction at its center.
Umberto Mezzadra spent years building PlayPauseBe - a brand whose core promise was mindful, screen-free yoga practice. Physical card decks. A $200k Kickstarter launch. An audience that trusted him precisely because he told them to put their phones down.
Then he built an app.
That app, Floga, pre-launched in May 2025. It did $120k in revenue in 24 hours - entirely outside the app stores, so no platform commission was taken. It now sits at roughly $10k in monthly recurring revenue with around 4,000 active users.
Here is how that happened, and what the path actually looked like.
A Non-Linear Path to Product
Umberto studied economics, not engineering. He had a brief corporate stint, raised seed funding for a startup in 2012 that ultimately failed, and then stepped away from tech entirely. He became a ski instructor. Then a fashion photographer.
He came back around 2016 as a growth strategist, helping companies launch and scale. That phase was where he learned the mechanics of positioning, storytelling, and customer acquisition - not from a course, but from doing it in the field for other people's businesses.
In 2020, during the pandemic, he and his girlfriend launched PlayPauseBe as a side project: decks of yoga cards for practitioners who wanted to build their own sequences without a screen in front of them. The Kickstarter brought in over $200k in the first month.
A year and a half ago, he started wondering whether the same intention behind those physical products could translate into software - without losing the thing that made people trust the brand in the first place. One evening, he sketched a concept, found a developer in Lisbon, and began building Floga.
Validating Before Building
Before any code was written, Umberto spent time talking to people in his target market - but carefully. He never revealed he was building an app. He never led them toward an answer. He asked about problems, not solutions.
He tested with existing PlayPauseBe customers specifically because those people were already invested in the brand. Their feedback was genuine precisely because they did not know what he was looking for.
Once the core idea held up under honest questioning, he and his team set a constraint: what is the minimum state of development at which the app conveys enough value - present and future - that early adopters would pay for it? Not a finished product. Just enough substance and a clear direction.
The answer was two fully built yoga styles and a roadmap people could believe in.
Three Problems No One Warned Him About
Non-technical founders building mobile apps face a specific kind of learning curve, and Umberto walked straight into it.
Not knowing what "good enough" looks like. As someone without a development background, the hardest judgment call was knowing when to keep improving versus when to ship. That requires trusting your developer while still pushing on the things users actually care about.
Building an app for an audience that came to you because you told them to stop looking at screens. Floga could not follow the standard playbook of maximizing session length and notification frequency. The app had to support the practice and then get out of the way. That raised the design bar considerably - every screen needed to feel calm, intentional, and visually considered.
Scoping the MVP without gutting the vision. Cutting features you genuinely believe in is harder than it sounds. But if you do not, you never launch.
The design constraint, in particular, shaped everything. Staying true to the brand meant the product had to be aesthetically coherent from the first screen - not functional-first, beautiful-later.
The Stack
The technical choices reflect the priorities of a small team moving fast with limited infrastructure overhead:
Flutter for the app - one codebase for both iOS and Android, which eliminates the cost of building and maintaining two separate codebases. It also supports the polished, custom look the brand required.
Firebase for the backend - around $25 per month at current scale, handling authentication, database, and hosting without requiring custom infrastructure setup.
RevenueCat for monetization - manages cross-platform subscriptions and handled the lifetime deal at launch. Revenue analytics come from here too.
Vimeo for video hosting - clean playback, no ads, reliable performance.
OneSignal for push notifications.
Simple, well-documented tools throughout. AI coding tools work best with standard frameworks, and keeping the stack orthodox meant faster iteration.
How the Launch Actually Worked
Rather than a free trial or a subscription for a finished product, Umberto offered a lifetime deal for an app still under construction. To make that work, he spent weeks being completely transparent about what existed and what was still being built. He did not hide the incompleteness. He made it part of the pitch.
The entire launch ran outside the app stores, which meant Apple and Google did not take their standard commission. All revenue went directly to the company.
One tactical decision shaped the outcome significantly. After weeks of warm-up content, Umberto drew a clear line: no refunds, the deal was the deal. If someone had doubts, they could wait for the subscription tier later. But the lifetime offer would close. Combined with limited spots and a short window, that framing moved people from consideration into a decision.
"Perfectionism is often just fear dressed up as preparation. Real breakthroughs happened the moment we put something unfinished in front of real people and let their feedback guide us."
Two growth channels worked in parallel. The existing PlayPauseBe audience provided a warm starting point - people who already trusted the brand and just needed to understand why the app made sense coming from them. Simultaneously, he ran lead generation to reach people who had never encountered the brand before, using a content buffer of emails, videos, and landing pages to educate and build credibility before asking for a purchase.
What He Would Do Differently
Umberto's advice is practical, not philosophical:
Stop waiting for the product to be finished before charging. Monetizing early funds the build without giving up equity and forces you to get clear on what people actually value.
Build an audience before you build the product. Launching into warmth, to people who already trust you, is a different experience from launching into silence. Start teaching and sharing before you have anything to sell.
Do not underprice. Founders set low prices out of fear. It leaves money on the table and signals low value. Price intentionally.
Talk to real people, and do not lead them. Ask about their problems, not about your solution. Polite encouragement is not data.
Ship before it feels ready. The discomfort of putting something unfinished in front of people is minor compared to the cost of building in isolation.
Where Floga Is Headed
The near-term focus is onboarding and conversion - improving how new users move through the app and into paid plans. The longer-term vision is broader than a single app.
Umberto describes Floga as one part of a larger ecosystem: physical products, a yoga school, apparel, teacher certification, and community - all connected and all held to the same design and quality standard. The app feeds the school; the school feeds the app. Each product in the ecosystem reinforces the others.
Follow the build on Instagram at @flogaapp, where the team shares regular video updates from behind the scenes. The website, Floga.io, is currently mid-rebrand, so expect it to look different
Their first after-hours call was a $20,000 job.
Air Texas was paying $2,000 a month for an answering service that couldn't close jobs.
Their first after-hours call with Podium’s AI Employee booked a $20,000 job.
Now no call goes unanswered after 5PM.
