The work it expands
The agent handles exploratory and recurring questions that often sit between a dashboard and a request to the data team
- Compare periods and segments
- Investigate changes and possible causes
- Prepare a readout for a decision
Have questions? Call us on +34 606 58 93 56
AI data analyst
An AI analyst that works with your business rules
Rumbleo answers questions over internal data in natural language, compares results, and keeps investigating in the same conversation. The data team retains definitions and control
Questions, clarifications, and follow-ups in one conversation
Metrics and definitions shared across the company
Read-only access with permissions respected in every answer
Revenue is flat, profit is down: investigating margin with AI
A worked example of investigating a margin decline, comparing channels and checking missing evidence before attributing a cause.
RumbleoExample using fictional data
Revenue is flat. Why has our margin fallen?
Revenue is unchanged at €100,000. Gross margin fell from 35% to 30%. The decline is concentrated in the online store.
| Channel | Revenue | Previous | Current |
|---|---|---|---|
| Online store | €50,000 | 40% | 32% |
| Stores | €50,000 | 30% | 28% |
| Total | €100,000 | 35% | 30% |
Gross margin = (net revenue − product cost) / net revenue. Like-for-like comparison periods.
Question 1 of 3: Revenue is flat. Why has our margin fallen?
What it needs
A reliable answer starts with shared definitions and a data connection prepared by the organization
- Documented metrics and dimensions
- Data the user is already allowed to query
- Enough context or a clarification from the user
What stays controlled
The agent can investigate without changing permissions or taking control of the database
- Identity and permissions
- Volume and time limits
- Usage monitoring and records
Questions suited to an AI data analyst
It is especially useful when a question needs comparison, follow-up, and an explanation beyond one number
Finance
“Why did margin fall versus last month?”
Break down the variance by business unit, category, or component and validate hypotheses in sequence
Operations
“Where is the increase in incidents concentrated?”
Compare sites, suppliers, products, or periods through the dimensions available in the model
Growth
“Which cohorts explain the retention change?”
Explore acquisition, behavior, and change over time without requesting a new view for every cut
Continue evaluating Rumbleo
Frequently asked questions
Does an AI data analyst replace the data team?
No. It reduces repetitive questions and expands access to analysis while the data team continues to define metrics and protect their quality
Can it invent a metric?
It should not. Rumbleo works with company-approved definitions and asks for context when a question has more than one valid interpretation
Do users need to know SQL?
No. People ask in natural language and receive the analysis without writing or reviewing technical queries
A demo built around a decision
Test a real analytical use case
Bring a question that currently waits for the data team and we will explore how to answer it with your metrics and ways of working
We will start with a real question from your team