Our first product Sunflower never should have existed.
When we were raising money for it, every VC told us “I hate this idea idea, but I love you as a founder Koby.”
And honestly they were right for the most part.
To build Sunflower in 2024, before Opus 4.5, you basically needed to hire 5 software engineers, a product designer, and a data lead, and then let them build for over a year. Over $1M was needed to create this product.
In December of 2025 we used Opus 4.5 to create Cashed Out, “Sunflower but for gambling”, in less than 48 hours.
It didn’t just have the same product quality, it was BETTER than Sunflower for gambling addicts.
Higher D1. Higher D7. Almost double the initial CVR into free trials. Better reviews “This product saved my life, whoever built it deserves to be a millionaire.”
We built a SMALLER product, and it was BETTER for a core segment of users than our bigger swing. This planted a seed in our minds that took root and began to grow like a weed no matter how hard we tried to ignore it.
Conventional VC wisdom has always been “build one product”. Consumer VCs are looking for the next billion dollar app. But our users and data consistently has been telling us about a path in a different direction.
Thesis: AI has opened up an ocean of “smaller” more personalized app experiences.
It never used to make sense to invest $1M into an app that would only make $500k. But now a single software engineer working together with AI Agents can make that same $500k app in a weekend.
We thought maybe Cashed Out was a fluke.
So we launched Grace. Then we launched Amanah. Then we launched Baseline. Then we launched Daylight. Then we launched New Parent.
App after app after app our users were telling us the same thing.
The real magical PMF was the fact we were building smaller pieces of software for use cases that could never have “deserved” a product before. We were building $10M apps (by 2024 standards) for niches that would only support $500k of revenue.
You can think of this like drilling for oil. AI has created a 100x improvement in efficiency that is allowing us to drill for oil that never used to be profitable to capture.
Atlantis is our platform for AI empowered workflows to fully ship apps from idea to MVP to the 1000th iteration and GTM fully baked in.
We leverage AI to ship value directly to our end customer. A single software engineer is capable of driving $2M+ in value per year. Atlantis is our platform that gives us leverage and scale so we are capable of building not just 10 apps, but 10,000.
We believe that there is $1T of new apps just waiting to be built. An ocean of new experiences has now been unlocked.