01
- Rescue
- GenServers timing out, LiveView sessions dropping, releases that will not boot. We stabilize production before investigating what the supervisor was hiding.
02
- Phoenix and LiveView
- LiveView mount and event flow, Channels at scale, PubSub topologies and the asset pipeline tuned for Phoenix 1.7 and up.
03
- Supervision and OTP design
- Supervision trees reviewed, state rebuilt from the right sources, GenServer boundaries drawn so no single process becomes the bottleneck for the whole app.
04
- Ecto and database work
- Pool sizing against real traffic, query performance, constraint-aware changesets, migrations that run safely on a replica-promoted production.
05
- Clustering and deploys
- libcluster topologies, mix releases, rolling deploys that drain cleanly, and distributed Erlang set up so cross-node calls do not fail under partition.
06
- Managed Tech
- Steady work on Elixir and OTP upgrades, Hex dependency updates, telemetry dashboards, and the weekly pass that keeps supervision trees matching what the application actually needs.