Managed databases
Provision a Postgres or Redis instance with a connection string in seconds — persistent, with a volume and a dedicated public port. No separate infrastructure to run.
Provision
The dashboard is the quickest way: Deploy → Databases → Provision. Pick an engine, name it, and copy the connection string.
Over the REST API:
bash
curl -X POST https://api.ai2in.dev/v1/databases \
-H "Authorization: Bearer ai2in_live_…" \
-H "Content-Type: application/json" \
-d '{"engine": "postgres", "name": "app-db"}'Response includes a ready-to-use connection string:
json
{
"id": "…",
"engine": "postgres",
"status": "live",
"url": "postgres://ai2in:<password>@<host>:<port>/app_db"
}Redis returns redis://:<password>@<host>:<port>.
Connect
Use the url from any client:
python
import psycopg2
conn = psycopg2.connect("postgres://ai2in:…@…:15432/app_db")Notes
- Instances are team-scoped; the connection secret is visible only to your team.
- They persist independently of sandboxes — delete them explicitly from the dashboard or
DELETE /v1/databases/:id. - Trusted first-party images run natively (not under gVisor) for full database compatibility.