# DataQube > DataQube is an enterprise AI data analyst that runs entirely inside the customer's own infrastructure — on-premise, private cloud, or fully air-gapped. Teams ask questions in natural language, run governed analysis against approved databases, files, and tools, and turn the results into auditable threads, notebooks, workflows, and presentations. Customer data never leaves the customer's perimeter, and every important answer is traceable to the query, file, model run, or permission decision behind it. DataQube is built for regulated industries, especially financial services (asset management, banking, insurance). It is licensed by concurrent active sessions rather than per seat, and licensing works fully offline. ## Product - [Platform overview](https://data-qube.com/product): Threads and notebooks, institutional memory, scheduled agents running deterministic workflows, governed presentations, a tool catalog, and audit-grade permissions — one governed system. - [Security & trust](https://data-qube.com/security): Deployment boundaries, per-user isolation, immutable audit, and SOC 2-aligned controls, designed for air-gapped operation. - [Pricing](https://data-qube.com/pricing): Concurrent-session licensing with Pilot, Enterprise, and Sovereign tiers. ## Industries - [Industries overview](https://data-qube.com/industries): Governed AI analysis for regulated sectors. - [Asset management](https://data-qube.com/industries/asset-management) - [Banking](https://data-qube.com/industries/banking) - [Insurance](https://data-qube.com/industries/insurance) ## Resources - [Blog](https://data-qube.com/blog): Engineering and product writing on governed, air-gap-capable AI analysis. - [Customers](https://data-qube.com/customers): Representative workflows for regulated institutions. - [Contact](https://data-qube.com/contact): Book a demo. - [Blog feed](https://data-qube.com/blog/feed.xml): Atom feed. ## Blog - [An AI notebook for power users: one kernel, two sets of hands](https://data-qube.com/blog/ai-notebook-for-power-users.md): Most AI analysis tools treat everyone as a business user — answers in, questions out, experts exiled to another tool. DataQube is a shared workbench: the agent and your quants co-work in the same kernel, on the same dataframes, onto the same record. (2026-07-21, Product Engineering) - [Deterministic AI workflows: a language, not a recording](https://data-qube.com/blog/deterministic-workflows.md): Deterministic AI workflows can't be built by saving a session as a job — repetition needs properties you can read before anything runs. Why DataQube workflows are written in a language: ordered steps, typed parameters, and a grant manifest that can't lie about what it touches. (2026-07-08, Product Engineering) - [Context compaction without amnesia](https://data-qube.com/blog/compaction-without-amnesia.md): Every long AI investigation outgrows the context window — the honest question is what happens then. How DataQube separates the model's working context from the append-only record, makes compaction a visible event, and keeps compacted threads fully auditable. (2026-06-18, Engineering) - [AI agent permissions: why ours asks first](https://data-qube.com/blog/ai-agent-permissions.md): AI agent permissions are the question under every enterprise deployment: what can it do without a human, and who decided? How DataQube treats capability as a grant, stops mid-run to ask, and puts every denial on the record. (2026-06-02, Product Engineering) - [Air-gapped AI agents: what zero egress actually takes](https://data-qube.com/blog/air-gapped-ai-agents.md): Air-gapped AI deployment is not a feature flag — it is a constraint you either designed for or didn't. Where the model runs (your cloud with zero data retention, or fully inside), how delivery and licensing work with no route out, and what support means when nobody can SSH in. (2026-05-12, Platform Engineering) - [Your AI analyst needs a brain, not a vector store](https://data-qube.com/blog/memory-graph-for-agents.md): AI agent memory built on similarity search retrieves what sounds related. Institutional knowledge is structural — what applies where, what supersedes what, who validated it. Why DataQube's memory is a knowledge graph with supersession semantics, and how it knows when it's wrong. (2026-04-22, Research) - [Provenance by construction: AI analysis your auditors can replay](https://data-qube.com/blog/provenance-by-construction.md): An AI audit trail works only if it is built into how answers are produced, not written up afterwards. How DataQube constructs provenance — queries, citations, workflow runs, refusals — so a reviewer can replay any figure without system access. (2026-03-30, Engineering) ## Optional - [German site](https://data-qube.com/de) - [Spanish site](https://data-qube.com/es) - [French site](https://data-qube.com/fr)