We modernised and stabilised a heavily customised Mastodon-based social platform on Ruby on Rails — without disrupting live community traffic.
A client running a highly customised Mastodon instance needed long-term engineering support. The platform had drifted far from stock Mastodon, so day-to-day reliability depended on deep Rails knowledge plus the ability to debug across the full stack — application code, background jobs, and production infrastructure.
Customisations had introduced hard-to-reproduce failures across Rails logic, React integration points, and supporting services (PostgreSQL, Redis, Nginx). Feature work and bug fixes had to land without taking the live community offline, and every change needed careful troubleshooting because the codebase no longer behaved like a standard Mastodon install.
Our team embedded into the project and took end-to-end ownership of stability and delivery. We diagnosed production issues across Rails and React, tightened the surrounding stack, and shipped incremental improvements so the platform stayed maintainable under real usage — turning a fragile custom fork into something the team could keep evolving confidently.
