Your protein_backend service's logs show that API requests (particularly /api/me) have been timing out at ~1 minute since the platform-wide service disruption that began at 02:25 UTC today. The backend was redeployed at 07:16 UTC but the timeouts persist, which suggests the Postgres database connection is unhealthy - your Postgres service hasn't been redeployed since May 9. Please trigger a redeploy of your Postgres service from the dashboard, then redeploy protein_backend afterward so it picks up a fresh connection.