Stack · Infra
Kubernetes
Zero-downtime deployments and migrations, rolling updates and scaling.
2 places3 challenges
My experience
Filetable runs on a Kubernetes cluster in Switzerland: deploys in the middle of the day, two versions side by side, workers added on demand. Before that, CYSEC's services.
Where I used it
- Since 2026
Filetable· CTO & co-founder - 2018–2020
CYSEC· Lead Software Engineer
The challenges
- Launch fast, then last: SaaS, AI, and the moment to own your stackVercel, Supabase, Trigger.dev and AI let you launch a product in a weekend. But there comes a point where running your own stack costs less, and where you need an engineer in charge.8 minEssays
- Slow queries: from symptom to cause, without guessingMeasuring every query without slowing the app down, recognising three shapes, and four real cases where the culprit was not the one we expected.5 min
Filetable - Deploying in the middle of the day: two versions, one database, open tabsFor a few minutes, the old and the new version run side by side on the same database. And an open tab can stay on the old one for days.6 min
Filetable
In the same family
DockerDockerIdentical environments from laptop to production, reproducible CI builds.Used at
Filetable
Branchlate
CYSECSee the details →TemporalTemporalDurable workflows for AI and long-running jobs: retries, priority queues, open source and self-hosted.Used at
FiletableSee the details →OpenTelemetryOpenTelemetryTraces, metrics and slow queries: knowing why it is slow before guessing.Used at
FiletableSee the details →GitLab CIGitLab CIFull CI/CD: tests, builds and automated deployments on every merge.Used at
OneDoc
CYSEC
ID QuantiqueSee the details →NginxNginxReverse proxy and TLS in front of SaaS services.Used at
CYSECSee the details →LinuxLinuxServers, scripting and deployment automation.Used at
Pictet
CYSECSee the details →