/0.4 · Ask

Your data, ready for your AI.

Work with your data in Claude, ChatGPT or your own agents. Ask questions, run analysis and build dashboards in Dataloader or as Claude artifacts, all from the same model, through the Dataloader MCP server and API.

What it is

  1. 01

    People use dashboards. Claude, ChatGPT and your own agents use the MCP server. Both read the same model, so they give the same numbers.

  2. 02

    Answers come with their source, ingest run and time.

  3. 03

    KPIs without enough connected data are rated low confidence or not shown, instead of guessed.

The difference

Let the AI ask. Let Dataloader calculate.

✕

The AI calculates

You paste raw data into a language model and ask for metrics. The model guesses the arithmetic and often gets it wrong.

~

The AI writes code

The language model writes SQL or Python to query your database. Different prompts produce different logic, risking wrong joins or double counts.

✓

Dataloader calculates

Defined KPIs are calculated on linked data by the Dataloader server. The AI reads the result with its source, so you get the same answer every time.

Language models are good at language, not arithmetic.

Capabilities

MCP and API for supported AI tools.

Dashboards in Dataloader or as artifacts

Ask for a dashboard in plain language. Publish it in Dataloader for your team, or keep it as a Claude artifact you can share. Both read the same model.

MCP server

Claude, ChatGPT and your own agents read the Observatory model through MCP.

Provenance in every answer

Values come back with their source, ingest run and time, so an answer can be checked.

EU language-model policy

Language models run with providers in the EU or EU-hosted regions, receive only redacted, pseudonymised data, and are contractually barred from training on your data. Generated values are tagged so they stay distinguishable from observed ones.

In practice

What Ask did for customers.

Illustrative example

ROAS × margin

Marketing that follows profit

The question
“Which campaigns actually make money?”
What the platform did
Ad spend from Google Ads and Meta Ads meets order revenue and cost prices in the model, so an assistant can rank campaigns by margin, not only by revenue.
Result
A question your team can ask in Claude or ChatGPT, answered from the same data as the dashboard.

Illustrative example

Illustrative example

Unit A: 31%

Gross margin, with its source

The question
“What was our gross margin by business unit last quarter?”
What the platform did
Sales come from the ERP and cost from accounting. Where purchase cost is missing, no figure is shown until it is connected.
Result
Every number links to its source.

Illustrative example

Customer result

+0.5 pp

Site-wide conversion rate

The question
“Can shoppers get answers on the product page?”
What the platform did
The product advisor answers shoppers on the product pages from the same model, with enriched product data behind it.
Result
From properly enriched product data and product advisor agents on the product pages.

Owner-reported customer result

How it works

  1. 01

    Connect AI

    Point Claude, ChatGPT or your own agent to the Dataloader MCP endpoint.

  2. 02

    Ask

    Ask in plain language. The assistant reads the Observatory model.

  3. 03

    Check

    The Observatory returns values with their source, ingest run and time.

Technical notes

For your IT and data team.

Provenance
Queries return data with its original provenance tags.
Access
Read-only access wherever the provider allows it.
MCP
MCP server implementation for supported clients.
AI models
Works with Claude, ChatGPT and other MCP-compatible clients.
Sources
Reads the Observatory model.

Questions

Frequently asked about Ask.

Why not let ChatGPT calculate our numbers?

Language models cannot reliably perform arithmetic on large datasets. With Dataloader, metrics are calculated on linked data by the server. The AI simply reads the result with its source, so you get the same answer every time, with its source.

Can we build our own dashboards?

Yes. Ask Claude, ChatGPT or your own agent for a dashboard in plain language. Publish it in Dataloader for your team, or keep it as a Claude artifact. Every number comes from the same model, with its source.

Which AI tools work with Dataloader?

Claude, ChatGPT and your own agents, through the MCP server and the API.

Will my data be used to train ChatGPT?

No. Language-model providers used by Dataloader are contractually barred from training on your data.

Do we still need dashboards?

If you want them. Dashboards and AI assistants read the same model, so they give the same numbers.

How does the AI know where a number comes from?

Every value in the model carries its provenance: source, ingest run, time, and who produced it. Answers return it with the value.

Where do the language models run?

Language models run with providers in the EU or EU-hosted regions, receive only redacted, pseudonymised data, and are contractually barred from training on your data. Generated values are tagged so they stay distinguishable from observed ones.

See what your data can answer.

Talk to us