**#project-sahel-noix** -- Slack **Aminata Kone** 8:14 AM We need to talk about our deployment process. Two production outages this season. March 15 -- 3 hours down, 12 tonnes untracked. April 2 -- 45 minutes, during a shipment inspection. Both caused by code pushes. Our developers push directly to production. No testing, no staging, no checks. I've been told a CI/CD pipeline would fix this. I've also been told there's a way to connect our developers' AI tools directly to the database so they stop asking me what the tables look like every time they start working. I need this solved before next season starts. We're scaling from 500 to 800 tonnes and I cannot have the tracking system going down during processing. What do you need from me to start?