The consultant leads discovery: it asks one thing at a time, adapts its depth to whoever is answering, and records each answer as a traceable requirement. You do not need to know what it needs to know.
The problem
A conversation with an AI tool starts with an empty box and the whole question on your shoulders. People outside the data field do not know what to declare, forget the volume, the team and the budget, and get a confident answer calibrated for a scenario that is not theirs.
It runs the interview like a senior consultant: asks what actually decides the architecture, challenges the request when the scenario does not support it, and records requirements and assumptions with the quote preserved. Recommendations come only after understanding.
How it works
The engagement is a state machine with a visible track: Interview, Analysis, Audit, Deliverables, Concluded. In the first phase it recommends nothing, on purpose: it asks about business, volume, latency, team and what already exists, and every answer becomes a labelled requirement.
With someone who knows nothing about data, it explains through analogies from their own business. With a specialist, it goes straight to the technical point. The depth follows whoever is in the conversation, instead of aiming at an average user who does not exist.
Now: alternatives compared against weighted criteria, cost computed per tool, assumptions on record.
We want to move everything to a data warehouse. Which one do you recommend?
Before choosing, I need two numbers I do not have yet: your active volume and how many people query per day. You mentioned 40 GB and 25 BI users. At that size, a warehouse usually costs more than it solves.
I will record that as an assumption, with the condition that would invalidate it: if active volume passes 1 TB, the math changes.
Makes sense. Compare the two options then.
Requirements captured
Assumption on record: "active volume stays below 1 TB". Re-checked on every read of the environment.
Part of the answers are not with you. The consultant builds a block of questions per role (finance, operations, engineering) and each person gets a public link to answer what they know, without creating an account.
The answers come back into the engagement as material from that role, and it is in the conversation that the consultant crosses them and points out where two areas contradict each other.
consulting interview
A few questions for the Finance role
Your answers go straight to the project’s data consultant. Answer with what you know: "I don’t know" is also a useful answer.
No account and no password: the link is the authorization, and it stays out of search indexes. Whatever comes back enters the engagement as that role’s answer.
Under the hood
Discovery works because of conduction rules in the product, and these are they.
What comes next
A public link per role: finance, operations and engineering answer on their own time, and the answer enters the engagement identified.
See it from the inside →DecisionDecision matrixWeighted criteria, justified scores, and each option declaring how much it costs and how long it takes to leave it within 18 months.
See it from the inside →MemoryMemory and assumptionsFacts, decisions and assumptions outlive the conversation, and each assumption is watched with the condition that would invalidate it.
See it from the inside →Paste the structure, get the most serious findings in about two minutes, and decide later whether an engagement is worth opening. No account, no card.