Greater AI autonomy requires rigorous verification.
We test how far AI can take a change, where it fails, and what evidence is needed to trust the result.
ButterOpsCase studies / ButterTrip
Our playground for learning what works with AI.
ButterTrip is the travel product we build and maintain across web, iOS, and Android. It gives us room to explore approaches your team may not have time to investigate.
Building + maintaining
First · an empty repository
We started from a blank codebase.
We chose the stack, set the standards, and built with AI. It let us explore how much experienced engineers could delegate, and how far someone without an engineering background could go with the right setup and checks.
Now · a substantial existing system
The learning continues through maintenance.
We keep changing what already works: testing new models, tools, and techniques, maintaining existing features, and making substantial architectural changes. We learn where approaches help, where they fail, and what it takes to use them responsibly.
New modelsTools + techniquesReal executionMaintenance
What that teaches us
Your team has a roadmap to deliver. We have a product where we can keep exploring.
We test new approaches through building and maintenance, learn their limits, and bring the useful lessons to your team.
The right setup lets more people build.
ButterTrip helped us learn how someone without an engineering background can build real features, with clear boundaries and reliable verification.
The product we’re building
A place for the whole trip, across three platforms.
Product design previews for web, iOS, and Android.
Follow what we’re building.
Visit ButterTrip. App Store and Google Play releases are coming.
Coming to theApp Store COMING TOGoogle Play
What we bring to your company
Experience your team can adapt to its own work.
We bring what we learn to your team, then test what fits your systems and your goals. A method that works for us still needs evidence in your environment.
Define what the change must preserve.
Turn requirements into checks that establish whether the new behavior is correct. At PLA, this work also uncovered long-standing bugs.
Run it against realistic systems and data.
Exercise the whole process in an environment that reflects the real application. At PLA, running the change found seven bugs that AI review had missed.
Check what happens when operations overlap.
Examine the rules for work happening at the same time, where tests and review can miss failures. At PLA, this exposed a bug in the existing system.
Measure reliability alongside speed.
At PLA, roughly 4× the engineering throughput came with reverts falling from 1.4% to 0.25%.
Free45 minutesWith Tyler and Julian
Let’s talk about what your team could do with AI.
A free, 45-minute conversation with both of us. Tell us what you’re trying to do and what keeps getting in the way. We’ll help you think through what to do next, whether or not you work with us.
On the call · 45 minutes
- 01
We learn how your company works and where AI could make a difference.
- 02
We sketch the scope of working together, right there on the call.
- 03
We agree on next steps together, and nothing is decided for you.
After the call · 24 to 72 hours
- 04
If it makes sense, you get a written proposal within 24 to 72 hours.