We Didn't Mean to Build an AI Automation Engine: The Origins of B3OS

Many of the best products we use today started as fun experiments, often peripheral to the team's main focus. Slack started as a browser game - the game failed, but the internal messenger grew to become a $27B company. Facebook began as a quirky (controversial) experiment at college that evolved to revolutionize our entire social world. Across so many successful startups, experimentation and play leads founders to uncover PMF.

When Glitch failed to gain a sustainable audience, the team realized their internal chat tool, designed for team collaboration across different cities, was more valuable than the game itself

The origin story of B3OS is similarly rooted in experimentation and play.

B3 started in mid-2024 and for the first year and a half our sole focus was building and shipping high quality, onchain consumer products, games & infrastructure. Onchain gaming was already taboo, but we believed that we could change the narrative by creating novel experiences for regular people that truly belonged onchain. Our motto, "you can't try the same things and expect different results," was taken to heart - maybe too much. In 12 months, we launched 14 products, each built on a new set of blockchain primitives, all providing a distinctive experience for users.

Many of these products saw strong results initially - most notably Basement, our gaming gaming platform, saw over 200k wallets and 12m transactions in just a few months. We debated investing more in customer acquisition, but the reality was none of these products had PMF. Nonetheless, we had inadvertently built some pretty remarkable tech at the foundation of each of the 14 products.

Building for consumers in gaming requires simplicity, reliability, interoperability, and a native way to interact with web2 - all massive gaps in current crypto infra. The workarounds most crypto enthusiasts deal with are simply not good enough for the masses. Solving for these missing pieces was nothing short of a nightmare. If it was just one small thing to fix or some slight inconvenience to work around, B3OS may have never been created. In hindsight, the broken state of crypto was a blessing: sure we uncovered gaps that took us countless eng hours to find solutions for, but those solutions became the formative lego building blocks of what we call B3OS.

Around 10 months ago, B3OS began. Initially B3OS was Anyspend 2.0, an expansion of our payments API that allows anyone to onramp and spend any token on any chain. The idea was, we are building many products that require nuanced payment systems, why don't we optimize Anyspend to help us build these products faster and more reliably. Building modularly, we kept adding hooks and plugging in primitives and audited smart contracts from existing products. We started being able to use the product to improve our business. We built products like Walletgas (that would have taken days and multiple devs), and forked our own versions of existing SaaS subscriptions that saved us thousands of dollars a month. BAM! We had PMF... or at least it solved real problems for ourselves.

Excited, we began to pressure test the product with other teams and builders. The value proposition was consistent - build faster, more accessibly, and more reliably - but each team, builder, and trader had a different workflow they wanted to solve for. That is the beauty of B3OS: every stack is different, and requires a bespoke and reliable solutions. Without B3OS, teams are forced to piece together expensive, disconnected services or build for themselves using LLMs that lack context on crypto and produce highly variable results.

B3OS enables builders to create custom, reliable, onchain workflows that save time and money - and even innovate new products and strategies altogether.

These conversations led to a number of day-one partners who will have early access to B3OS and help us build out public templates.

B3 is not one product - B3 is the team. We are veterans who built Coinbase Wallet, Base, USDC and much more. Our experience building onchain has afforded us unrivaled knowledge of crypto's strengths, shortcomings, and potential. Neither AI nor automation tools understand crypto like we do - so we built an AI automation engine that does.

More on the technical details, our vision, and soon-to-come GTM - but one more note on the way out:

B3OS solves real problems for builders, businesses, and traders - and has limitless potential to improve building, automating, and trading onchain.

If you want to chat B3OS please reach out to [email protected]

❤ Viktoriya