Data where it lives. The answer, in chat.

    Etedata joins sources through Corebrain. It does not replace a warehouse when you need one. It stops you building one only so you can ask a question.

    The last decade’s reflex was to copy. If sales and support must be joined, you launch an ETL, wait for the model and declare a “single source of truth” that is dirty a quarter later. In banking and in any firm that cannot take data out of perimeter, that reflex is expensive and sometimes de facto unlawful. Etedata starts from the opposite assumption: the stores already exist; the work is to orchestrate access, analyse and converse. Corebrain is the brain. The lake, if needed, comes later and with a reason.

    What a warehouse actually solves

    A well-built store serves long history, heavy joins and a team that lives on models. It makes sense when volume, retention or data science demand it. It does not make sense as the first step so leadership can ask “what happened this quarter” across three Postgres databases.

    The hidden cost is not the licence. It is the team that keeps pipelines, the latency until data is “BI-ready” and the copy you must protect as if it were production — because it is. Under an EU framework, every copy is surface.

    What querying at the source solves

    Less data movement. Less shadow. A layer that discovers schema, translates the question and runs with encrypted credentials. Etedata adds boards and chat. Corebrain can be embedded in your product via SDK. Ask Intelligence is the question-answer face for the business.

    The first connection, in a clean case, is measured in minutes, not an integration program. That does not remove the work of giving the schema business names. It removes the toll of “lake first, then we talk”.

    Banking and sovereignty, without the slogan

    In banking environments data has residency, retention and an auditor who will ask where the query was processed. “We sent it to a model in another cloud” is not an architecture. It is a deferred incident. Corebrain’s design — data at source, schema context, controlled execution — exists for that conversation, not for an LLM fad.

    If your supervisor or DPO cannot explain the flow, you do not ship it. The diagnostic is there to see whether the current perimeter can stand an NL layer or whether access must be closed first.

    A practical rule for choosing

    Does the question fit the stores you already run, and is the join occasional? Source. Do you need five-year history, heavy models and an analytics team that publishes datasets? Warehouse — and Etedata can still be the conversation layer on top if the warehouse is one more source.

    Are you copying “in case the business asks”? Stop. That sentence funds eternal ETLs. Ask first. Copy when the question shows up twice a week and the source cannot take it.

    How it enters Orbit

    Etedata and Corebrain are not a separate project if a program is already running. They activate when the cycle needs them: same credentials, same owner, no “data platform” RFP. Onboarding maps sources, permissions and the ten questions that actually keep you up.

    If there is no Orbit, the product can still be scoped. Delivery does not change: diagnostic, connection, validation, deploy. What we do not do is a six-month “eternal POC” to prove SELECT works.

    Choose with craft
    QuestionSource (Etedata / Corebrain)Warehouse / ETL
    Time to first answerDays, sometimes hoursWeeks or quarters
    Copy of the dataMinimal; queried where it livesCentral; must be protected
    Occasional join of two storesNatural caseOversized
    Long history / data scienceNot the designNatural case
    Surface for an auditorFewer places to explainMore copies, more contracts
    Required teamNo data department requiredYes, if the lake is serious
    Questions

    Lake or brain.

    Does Etedata replace Tableau or Power BI?

    Not necessarily. It covers analytics and chat over connected sources. If you already have BI that works for fixed dashboards, we do not rip it out on dogma.

    Which sources ship natively?

    PostgreSQL, MySQL, MongoDB, Redis, SQLite and REST APIs. Others are evaluated. Data stays at the source.

    Can we embed it in our product?

    Yes. Corebrain is an SDK (Python and JavaScript) and an orchestration layer. Ask Intelligence and Etedata are Globodain’s faces of the same engine.

    Is Orbit required?

    No. Orbit is how many companies hire us monthly and pull data in when the cycle needs it. The product can be activated on its own scope.

    Let's put your business in orbit.

    Initial audit, monthly plan and clear cadence to adapt your technology to what the framework requires. Proposal in 48 business hours.