Modern data stack, in your account
The stack a ten-person data team would take a year to assemble, deployed into your cloud account with agents doing the pipeline work and your people approving through a governance gate. Nothing hosted by us; nothing to migrate off later.
Where you are
Getting from spreadsheets and a vendor warehouse to a governed, AI-ready data platform takes a team you cannot hire and a year you do not have. Hosted platforms shorten it but hold your data, lock in one engine and one AI vendor, and charge per seat forever.
How the engagement runs
Deployed in your account in a week
Ingestion, Airflow, dbt, Iceberg storage, a catalog, BI and Text-to-SQL, behind one console, inside your VPC with your IAM and your keys. We run nothing on your behalf.
Agents build, people approve
A pipeline agent plans the dbt models from a source schema or a request, writes them, runs them and repairs failures. Every plan and every diff goes through a governance gate your team approves; every action is logged.
Governance produced by the build
Column lineage, tests and personal-data flags come out of every dbt build, not from a separate catalogue project.
Any model, by policy
Bedrock, Anthropic, OpenAI or a local model, chosen per task by a data-residency policy you set.
Already built
- ●Our own estate runs this way and is demonstrated on platform.datap.ai: pipelines, dbt, lineage, BI and Text-to-SQL in one console.
- ●Text-to-SQL on Databricks and Snowflake over governed tables.
- ●Open source components throughout: Airflow, dbt, Apache Iceberg, Lightdash, OpenLIT.
What you bring
- ●A cloud account and VPC (AWS first; Azure and GCP mapped)
- ●An object store bucket and an IdP for SSO
- ●Read-only access to a few source tables to start