Evidence before confidence
Answers should resolve to real queries, documents, graph paths or other produced artifacts. If a claim cannot be supported, it should be dropped or clearly marked as unconfirmed.
About Datapane
What we believe — Responsible AI
Datapane.ai is an agentic intelligence platform connecting your data estate with the business context that makes it useful: definitions, BI reports, documents and human decisions.
Our Principle
Never ask anyone to trust “the AI said so.”
Every AI output should make its basis visible: the evidence it used, the governance that applied and the uncertainty that remains.
An honest answer with limits is more useful than a confident answer without proof. When the platform cannot verify a claim or compute a metric, it should say so.
Our Vision
One governed meaning layer for people and agents.
We see a future where enterprise data, business definitions and human decisions do not live in separate worlds. Every analytical and agentic workflow should be able to use the same approved context.
Connected sourcesGoverned meaningPeople & agents
Our Mission
Turn enterprise context into responsible action.
Our mission is to connect catalogs, data, BI definitions, documents and steward decisions so teams can ask complex questions and inspect how the answer was reached.
We build the context, trust and agent layers that help organisations move from scattered information to governed analysis, pipelines and decisions—without taking people out of the loop.
Our Values
How responsible intelligence behaves.
These values translate the deck’s trust, control and engineering principles into the standards we use to shape the platform.
Human authority
A steward’s explicit decision outranks an AI inference. Conflicts should show both positions, preserve the ruling and apply that governed meaning consistently in future work.
Governance by design
Authorization and business rules belong at the server and source boundary. Every person and agent should operate with a scoped identity and the least privilege required.
Honest fallbacks
When a model, search service or graph cannot complete its work, the fallback should be visible. Missing information is not permission to fabricate certainty.
Team
The people building responsible AI.
Meet the people shaping a governed, evidence-led way for organisations to work with data and AI.
Co-Founder & Managing Director
Asutosh M.
VP of Operations & Commercial
Yashar Rahmatollahi
Director, Engineering Administration
Bikash Rautraya
VP Of Customer Success
Prabhat Kumar Goyal
Independent Advisor, AI security & Global Impact
Dr. Tapapriya Mahanta
FAQ
Questions about how we work.
What does Responsible AI mean at Datapane?
It means AI outputs remain inspectable: the evidence, source, model or fallback behind an answer should be visible, and uncertainty should be stated rather than hidden.
How does Datapane keep people in control?
Agents can propose interpretations and actions, but governed decisions remain reviewable. Human assertions take priority over machine inference, and higher-risk work can require approval before it proceeds.
Does Datapane replace existing data catalogs and BI tools?
The platform is designed to connect the systems an organisation already uses. Catalog, lineage, BI definitions, documents and approved business context can contribute to a shared meaning layer rather than becoming isolated sources of truth.
How can I learn about careers or the team?
Open roles are published on the Careers page when available. Team profiles on this page come from Datapane’s managed company content.
Let’s connect
Start with your data and your context.
Tell us about your sources, BI tools and the decisions your teams need to make.
Talk to our team