7:50 · Music video
The Iceberg Open Lakehouse
The governed, portable data foundation agents need before analytics can become truly agentic.
Open videoFor analytics leaders, heads of BI, and CDOs
Agentic analytics for analytics leaders
The case, the rollout, and how to tell if you are ready.
An AI agent can do an analyst's work: read a question, find the data, run the query, check the result, and try again. What decides whether it is right is not the model. It is whether the data underneath has one definition, one copy, and one place access is enforced. AI agents plus governed semantic data equals agentic analytics.
AI agentsPlan, call tools, check the result, try again.
Governed semantic dataOne definition of the metric, one place access is decided.
Agentic analyticsAn answer with its lineage attached, not a guess.
Tools for analytics leaders
Decide, choose, and roll out
Four working tools, in the order most teams need them. Free, with no sign-up. The assessment is scored in your browser.
- 1Readiness assessmentFifteen questions on definitions, access, data quality, interfaces, and team. Scored in your browser into a maturity level with next steps for each weak area.
- 2Use-case libraryFinance, sales and revenue operations, supply chain, customer support, and marketing: the questions people ask, the definitions an agent needs, and the governance risks.
- 3Evaluation checklistVendor-neutral RFP questions on semantics, governance, lock-in, accuracy, cost, and operations, with what a good answer contains. Print it or download the Markdown.
- 490-day rollout playbookFive phases across thirteen weeks, from picking a domain to deciding whether to expand, with owners, exit criteria, and the risks that stall each phase.
Featured films
Two anthems for an open data future
Music-driven explainers for the open lakehouse and the open agentic stack.
7:37 · Music video
Open the Stack
A musical manifesto for open agent profiles, interchangeable models, and portable execution.
Open videoEither half alone gets you a demo.
Agents without governed data are fluent and wrong, at machine speed, with nobody checking. Governed data without agents is a well-run warehouse that still answers questions in a week.
Put them together and the long tail of questions, the ones that were never worth an analyst's time, becomes answerable. That is the whole proposition, and most of the work is on the data side.
Start here
What the term means, how it differs from the dashboard era, and why the earlier attempt at this stalled.
- What is agentic analyticsAn AI agent doing the work of an analyst against governed data, with the authority to run queries and the obligation to show where the answer came from.
- From dashboards to questionsDashboards answer questions somebody anticipated. Agents answer the ones nobody did, which changes what the data layer has to support.
- Text to SQL and why it stallsGenerating SQL from a question was solved years ago. Generating the right SQL against a real schema was not, and the gap is definitional rather than technical.
Three layers, three failure modes
An agent is not a model
It is a model inside a loop, reaching outside itself through interfaces. Knowing which of the three is failing is most of debugging one.
- The agentic stackThree separable layers, model, harness, and open interfaces, each with its own failure mode. Most agent problems are misattributed across them.
- The modelWhat the reasoning layer actually contributes, what it cannot know, and why treating model choice as the whole architecture leads people astray.
- The harnessThe program that runs the loop. It decides context, tools, retries, memory, and when to stop, which is where most agent behaviour actually comes from.
- Open interfacesThe four contracts an analytics agent speaks through, why each should be a specification rather than a product, and what a closed one costs later.
The other half
Why an open lakehouse
Agents multiply readers, multiply questions, and multiply the temptation to make a private copy. An architecture where the data sits in open formats, with one catalog deciding access, gets better under that pressure. One where reading requires a particular vendor's engine gets more expensive.
Five Apache projects settle five separate questions. Each one is a specification with more than one implementation, which is the property that makes the stack assemblable at all. What each means for analytics leaders:
- Apache Parquet: Agent questions become many small queries; Parquet statistics are why each one reads a fraction of the data and costs less.
- Apache Iceberg: One copy of each table that every engine and agent can share, with history that lets you reproduce the answer an agent gave.
- Apache Polaris: One place where access is decided, so an agent cannot reach data its user cannot, whichever tool it runs in.
- Apache Arrow: Results move between engine, tool, and agent without conversion, which keeps latency low when agents ask many questions.
- Apache Ossie: Metric definitions you can take with you, so the semantic layer does not become the next lock-in.
Running it for real
The part that decides whether a pilot becomes a capability: identity, evaluation, cost, and the order to do the work in.
- Governance for agentsIdentity, scoped credentials, and audit, arranged so that an agent cannot reach what its user cannot reach, and so you can prove it afterwards.
- One copy, many readersThe architectural property that decides whether agentic analytics scales past a pilot, and the specific way copy sprawl produces disagreeing answers.
- Evaluation and trustHow to know whether an analytics agent is right, why accuracy on a benchmark is not the question, and what to measure instead.
- Cost and latencyAgentic analytics changes the shape of the workload from few large queries to many small ones, which moves where the money goes and what needs optimising.
- Portability and lock-inWhat it would actually cost to replace each component of an agentic analytics stack, and why the semantic layer is now the most likely trap.
- Rolling this out enterprise wideWhat separates an agent pilot that impresses a steering committee from a capability a whole organisation uses, and the order to do the work in.
Longer reads
Published pieces that carry parts of this argument in more depth.
- DefinitionWhat Is Agentic Analytics?The definition, and what separates an agent answering a question from a chatbot guessing at one.
- SemanticsThe Semantic Layer: The Definitive GuideWhy shared definitions became load bearing the moment machines started asking the questions.
- GovernanceApache Polaris: The Catalog Standard for Lakehouses and AIOne catalog that both engines and agents resolve against, with access decided in one place.