Building Clevis - From idea to first 100 users

Idea 💡

Our story began with a simple idea – making AI-powered app development accessible to everyone. In Clevis, you can combine steps like prompting ChatGPT, fetching data from APIs, and generating AI images. Plus, you can trigger these apps either through a user-friendly interface, on a set schedule or through an API call.

Validation 📧

Believe it or not, me and my co-founder didn't start building straight away. Our debut was a simple landing page describing the product with an email form. We shared it on HackerNews, and surprisingly, it caught the attention of some AI tool listing websites, gathering around 400 email addresses along the way.

Building out the Foundations 🔨

With our email list growing, it was time to get to work. Thankfully, we previously launched a SaaS boilerplate that we used for this project. This gave us a lot of the basics that we needed and we could focus on building the unique features for our product. The majority of the time was spent on polishing the app editor UI and the mechanics of the app runner.

Greek Inspiration 🌴🍻

As the European summer rolled in, we felt the need for a change of scenery to give our MVP the final push. So, we embarked on a workation in Greece for a week. By the end of our stay, we had a somewhat functional MVP. And to be honest, it wasn't all about work; we managed to enjoy a few poolside beers as well.

The Big Launch (?) 🛬

With the MVP complete, we unveiled Clevis by sending a link to our email list. The results... did not quite match our expectations. The open rate was surprisingly low, and we're still trying to uncover the reasons behind it. On the bright side, the users who did open the email showed a rather promising conversion rate.

One Hundred Users 💯

After moving past the less-than-exciting email list launch, we brainstormed fresh strategies. Since a significant chunk of our email signups had originally come from the AI tool listing sites, we decided to reconnect with them. We asked if they'd be willing to feature Clevis again, now that we were officially launched. Thankfully, they agreed to give Clevis a second chance, and this time around, we witnessed a notable influx of user sign-ups, propelling us past 100 users.

Bug Bashing 🐛

Bug reporting tools became our constant companions as we were constantly fixing any crashes. We reached out to users via Slack and email to make sure that they didn't get stuck building their apps.

To Be Continued 🌅

So far it has been a blast seeing all the cool apps our users build using Clevis. We are excited to continue to grow the product and help more people tap into the potential of AI.