The self-describing database
Your data
should explain itself.
The managed database built for AI agents. Your data carries its own context — what it means, where it came from, whether you can trust it. Agents get the right answer, not a plausible guess.
We'll only use your email to contact you about early access.
Your agent doesn’t say “I don’t know.”
An agent answering from your data writes valid SQL and wrong queries. The errors are rarely syntactic. They are semantic: the agent does not know what your tables mean, which one is canonical, or which keys are safe to join on.
The failures repeat in patterns:
A join fans out, and the total inflates.
A predicate matches nothing and returns a confident zero.
The pipeline died Saturday. The results still look clean.
Per-order and per-item math in one aggregate.
Nothing throws an error. The number comes back plausible and wrong. Someone puts it in a board deck.
Engineering the problem away is hard. Agents rarely find flaws in their own SQL. Across benchmarks, they catch only a few errors in a hundred.
The check has to come from outside the model.
What a self-describing table looks like.
2. The fix
When data explains itself, agents stop guessing.
In Rune, every table and transformation carries mandatory context — what it means, who owns it, how fresh it is, where it came from. The context is the schema itself. It can’t drift because it is the definition.
Connect any agent over MCP. It discovers what exists, understands what it means, and traces how it was computed. One source of truth. The right answer.
Select a card to explore
One definition. Everything follows.
3. The payoff
Fewer tools. Fewer tokens. Lower cost.
A database with cataloging, lineage tracking, quality monitoring, and transformation pipelines built in. No separate tools to configure, no integration code to maintain, no context drifting across five systems.
Transformations run inside the database — no ETL infrastructure, no data shuttling between systems.
Every agent query against an opaque database burns tokens guessing. Self-describing data means fewer steps, fewer retries, smaller context windows. Lower inference costs on every question.
4. The trust
Proven technology. Clear purpose.
Rune is built on Postgres — the most widely adopted open-source database in production, trusted by millions of developers for three decades. Your team already knows it. Your BI tools, your drivers, your psql sessions — they connect unchanged.
For analytical workloads, DuckDB runs embedded inside Rune. Columnar speed that rivals dedicated warehouses, without a separate system to manage.
Two engines the community already trusts. One system.
Learn why we’re building Rune →Request early access.
We’re onboarding a small number of design partners — data teams who feel the pain of ungoverned data and want to shape the product with us. If that’s you, we’d like to talk.
We respond to every request personally within a few days. Your email is used only to contact you about early access.