Last year, in March 2025 I got laid off from Atlassian after working for Loom for nearly 6 years. There I experienced working on a codebase from a fresh Series B up to acquisition on pretty much every part of the product, so I lived through a few pivots. Before that I worked for Atrium, and again, a lot of pivoting. Time after time, the problem I see is the same:
New feature → Repackaging (does this justify a price change?) → Experiment (A/B test the packaging/price) → Roll out → Distribution
The problem is that none of the codebases I've ever worked with included these capabilities, and yet they are incredibly useful for a business to move through the difficulties of their environment.
Time passes and experiments pile up, technical debt compounds and your codebase becomes a fire hazard.
My family needed to move back to Europe (Portugal) for personal reasons, and once we settled I decided to just focus on building Refract and dogfood my own product along the way.
I've actually created a little blog article about this called Why I'm Building Refract After 6 Years at Loom.
The blog is mostly going to be about software engineering best practices by the way.
Oh, and if you're interested about the project, please hit me up, I'll gladly show it to you!