Mirio, a retrospective

This is a short product retrospective on Mirio, the first incarnation of the company that David and I started.

Now that we’ve moved on to something new, I wanted to take some time to document some of the things we created over the course of a year with a very small team. David, Gerardo and I designed and built everything here.

Mirio 1.0

We started with the idea of building better tools for teaching and learning programming. Most programming courses are just hours of video. These videos are a huge pain to produce and not engaging for learners.

We wanted to create a new format for learning where instructors could just record their code (alongside video) and users could play it back, edit it, and run it in their own browser.

We hacked together an early prototype, applied to Y Combinator, and got in.

We also brainstormed a bunch of names, picked one, and I threw together a quick consumer-focused brand identity.

Other the next 2-3 months, we designed and built an MVP.

Creators could open up our web-based editor, hit record, and start coding. This would create a sharable, interactive recording of their code that they could share with anyone. Viewers could edit and run the code directly from the video and branch the code to create their own projects.

We also build a complete intro to HTML, CSS, and JS course using Mirio.

Viewers could also generate code and ask for help from an AI-powered assistant.

Building supply on the marketplace was hard. Recruiting creators to build on the platform was slow and creating our own content took time away from building and selling. So, we started to look into B2B models to see if there were opportunities to help developer-facing companies create better sales and marketing content.

A few weeks before demo day, this led us to pivot to the next iteration of Mirio, a demo builder for API companies. A week later, we signed our first customer.

Mirio 2.0

Over the next few months, we built a no-code platform to help sales and solutions teams create customized demos for their API products.

We created a no-code demo builder that enabled non-technical teams to build interfaces from scratch and connect them to their API.

We shipped a ton of features along the way, including a presentation builder for telling multi-demo stories:

Interactive click-through guided demos with live and mocked API calls:

AI-powered creation of working demo interfaces:

Automatic demo styling to match customer brands:

We started ramping up sales and began to see traction but it didn’t come as quickly as we hoped. We came to the conclusion that the market for API demos wasn’t going to sustain the type of business we wanted to build. More importantly, we realized that this problem we stumbled on wasn't one that we were passionate enough to commit years to solving.

So, around Thanksgiving, we began to consider new directions. We landed on an idea that I'd first written down in my notebook years ago. We're more motivated than ever and can't wait to bring it into the world. You can learn more at getversive.com