DecisionBox for Microsoft SQL Server - Connect DecisionBox to your MSSQL to validate findings

Hey PH, Selçuk here from DecisionBox, If you're on SQL Server or Azure SQL, the database that runs your app already holds the data worth analyzing. The hard part isn't getting to it. It's figuring out which questions are worth asking, getting validated answers, and doing it without disturbing the workload the database is already serving. That's what we built DecisionBox for. Point it at a SQL Server, a database, and a schema. The agent writes its own SQL, validates every finding against the data, and ships a ranked backlog of insights and recommendations. No prompting, no question-writing. Today we're adding Microsoft SQL Server to the list of supported warehouses — and unlike most things in this category, the provider is built for a database that's also serving your app. Read-only, scoped by a SQL Server login. Create a login (or a contained user on Azure SQL) with SELECT on the schemas and tables you want exposed, nothing more. The GRANT model is the boundary; the agent cannot reach anything it hasn't been granted. Safe to run on a production database. Small connection pool, capped at 5. Per-query timeout, defaults to 5 minutes. Row counts read from sys.dm_db_partition_stats — engine-maintained, no full table scan to count rows. TDS encryption on by default. No writes, ever — discovery queries are read-only SELECTs. Two ways in, both standard SQL Server: a username/password login, or a full sqlserver:// connection string for Azure SQL or anything that needs extra TDS parameters. If your security team has already approved how the rest of your stack talks to this database, they have already approved DecisionBox. The whole SQL Server provider is in the public repo, AGPL v3 — the auth flow, the INFORMATION_SCHEMA reads, the T-SQL the agent writes. Read it before you turn it on. Same agent already runs against BigQuery, Redshift, Snowflake, Postgres, and Databricks. If your stack moves, your DecisionBox install moves with it. Happy to dig into the SQL login setup, Azure SQL specifics, or anything else in the comments.