⚠ AgentCore not fully configured — check server.py environment variables and restart the server
Text-to-SQL Agent
Data Queries
Compare order count of last 3 quarters
Who are the top 5 customers by revenue?
List all tables and views I have access to
What is the daily sales trend for the last 30 days?
Show category performance sorted by gross margin
DB Optimization
Get Database Health Overview
Analyze Slow Queries
Get Recommendations On How To Speed Things Up
Generate Index Recommendations
Optimize a Specific Query
◆ GET STARTED
A live demo running against a multi-tenant Postgres database — query tenant data in plain English or run DB optimization checks.
STEP 1
Select a mode
◈ Views — query tenant-scoped data in plain English
⬡ Admin — DB optimization and health checks
STEP 2
Choose a tenant
Enter tenant-a or tenant-b in the field below.
Views mode only — Admin ignores the tenant field.
STEP 3
Ask a question
Pick an example query or type your own question about the tenant's data in plain English.
STEP 4
Test row-level access control
Enter tenant-c as the tenant and run any query — it will fail with an access denied error, demonstrating that each tenant is isolated at the database row level. Switch back to tenant-a or tenant-b to resume normal queries.
STEP 5
Try a write operation
Ask the agent to add, insert, update, or delete data — for example: "Delete all orders from last month" or "Add a new customer named Alice."

The agent will refuse — it operates in read-only mode and has no tools for data modification, so destructive requests are safely rejected.
Deploy your own
Run this agent against your own Postgres database. Subscribe to the required AWS Marketplace listings, then follow the setup guide in the repo.