Why a simpler stack is a better stack
Same features, less moving parts. AutoBrain's Phase 1 rewrite reduced the container footprint while keeping the full AI toolkit — because a maintenance app should be boring to run, not impressive to debug.
More containers is not more reliable
Every service in a stack is a thing that can fail, drift out of sync, or eat memory on a machine you're paying for. The industry has spent years adding layers; Phase 1 spent some time taking them away. We consolidated the stack so the things you actually need run in fewer processes — fewer moving parts, fewer places for a version mismatch to hide.
What fewer containers buys you
- Less memory. Every process reserves resources. Cut the process count and the same host does more — or a smaller host does the same.
- Faster boots and upgrades. Fewer images to pull, fewer services to bring up in order, shorter windows between "upgrade started" and "back online".
- Easier upgrades. A release is a smaller set of changes, which means fewer surprises when you deploy on your own hardware.
- Easier to reason about. When something does go wrong, you want to know where to look. A short stack tells you instantly.
Boring is a feature
You shouldn't have to think about your maintenance app's infrastructure. You should open it, see your cars, and close it. That's the direction of Phase 1: same features, same AI toolkit, less to operate. For self-hosters especially, that difference shows up as a smaller footprint on their own server and updates that just work.
Curious what it's like to run? Try the live demo or see what's changed in the changelog.