Frequently asked questions

Which engines will PillarMesh support first?

PostgreSQL and ClickHouse are the planned initial managed warehouse engines. Supported versions and operating boundaries will be published before availability.

Does PillarMesh support Spark?

No Spark support is being claimed. The intended audience may know Spark-style processing, so the docs use familiar ideas such as transformations, dependency graphs, and reconciliation without promising Spark as an engine.

Can PillarMesh write to any destination?

No. A destination must be inside the supported scope declared for an Integration Contract. PillarMesh will operate only declared, supported connectors and destinations, and it will not silently substitute a destination when a request is unsupported.

Is PillarMesh a scheduler?

No. A planned execution graph expresses bounded work for an accepted Integration Contract. It is not a general-purpose scheduler for arbitrary workflows.

Can AI approve or change a plan?

No.

AI may assist with interpretation, explanation, and candidate generation only; it is not authoritative for legality, planning, signing, execution semantics, reconciliation, evidence, recovery, or state transitions.

Can I use PillarMesh in production today?

No. PillarMesh is in development. There is no supported installation, public API, SDK, production SLA, or availability commitment yet.

What does working with a design partner mean?

Design-partner conversations are intended to test problem framing, operating boundaries, and the usefulness of planned evidence with hands-on data teams. They do not imply production access or a commercial commitment. Use the design-partner contact path on the main site to start a conversation.