{"type":"video","version":"1.0","html":"<iframe src=\"https://www.loom.com/embed/9200a5ab7f214aa28fbbf7845ed4523c\" frameborder=\"0\" width=\"1756\" height=\"1317\" webkitallowfullscreen mozallowfullscreen allowfullscreen></iframe>","height":1317,"width":1756,"provider_name":"Loom","provider_url":"https://www.loom.com","thumbnail_height":1317,"thumbnail_width":1756,"thumbnail_url":"https://cdn.loom.com/sessions/thumbnails/9200a5ab7f214aa28fbbf7845ed4523c-0beeb9b54625eaf5.gif","duration":660.417,"title":"Qovery Agent Task","description":"This Loom explains Covery’s agent cloud for running AI agents in production inside a team’s infrastructure rather than sending data and secrets elsewhere. It contrasts two options: exporting data to a third party or building and maintaining your own event receiver, scheduler, model wiring, and governance. The author demonstrates a “Build and Deployment Optimizer” agent that runs regularly, set to trigger every Monday morning, collects metrics, checks code, and proposes improvements via a pull request. The agent is configured with templates, model access (using Entropiq), Slack output webhook for a recap, and governance egress rules, then a PR link is shared after completion."}