Everyone gets an analyst. Nobody gets data they shouldn't see.
Ask in plain words and get the answer with its chart. Your data team gets its request queue back.
Private Preview. Launching Soon!

An AI-first analytics platform, built for governance and trust.
The whole idea, in under two minutes.
A question comes up. Today it joins a queue.
Someone needs a number before the meeting ends. It becomes a request to the data team, and the meeting moves on without it.

Ask it while it still matters.
Type the question in plain words. Qeries writes the query, runs it on your data under your own row rules, and answers with a result, a chart, an analysis or a recommendation. And dig further with follow-up questions.
Ask in English, German, French, Arabic, Hindi or Urdu, and get the answer back in your language. The screens themselves are in English.

See how the answer was worked out.
Each answer comes from a query on your own data, and the chart is drawn from what that query returns. Before you rely on a source, open its data quality report straight from the answer.


Find what moved the number.
Ask why a number changed, in a Deep dive. When the question compares two periods, Qeries breaks the change down by segment and shows each part's share of the move. The parts add up to the total, so you can check the sum, and a market that sold last period and not this one counts too.
The breakdown is arithmetic on the rows the query returned, not a model's reading of a table: the model writes the sentences around figures that are already settled. If the parts do not add up to the total, the answer says so.

Try the move before you make it.
In the What-if planner, move spend between groups within the same total, or press Suggest a plan. Every figure is computed from your own rows, and the basis is written under it. A move beyond anything your data has seen is marked as thin ice.
Five modes: Budget split, Pricing, Capacity, Drivers and Supply.

Send it to the people who act on it.
Save a plan, share it with named colleagues, and turn it into a recommendation. Export it as PDF or Excel. Schedule the board by email, and each person receives only their own rows.


Pick a specialist, or build your own.
Specialist analysts come ready for a discipline and stay off until your administrator turns them on. Or build your own: give it a role, its guardrails, and the data it works from. Either way, it answers under the permissions and row rules of the person asking.

One dashboard, two logins, different rows, same URL.
The CFO sees every market. The Gulf marketing manager opens the same link and sees only the Gulf. Answers, saved plans and scheduled emails follow the same rules.


Start from the data you already keep.
Upload a file or connect a database. Evaluations start on uploads; database connections open type by type.
Qeries connects read-only: it does not change the data in your sources. With Amazon Athena, it writes each query's output to a results location in your own AWS account.


Security, in three lines.
- Each viewer sees only the rows they are entitled to, enforced on the server. A source's owner and your admins see all of its rows, so give admin rights with care.
- Your data is stored in Frankfurt, in the EU. When you ask the AI a question, the question, a limited sample of result rows and any file you attach are processed by Anthropic outside the UAE. We do not offer UAE hosting today.
- Two-factor and Microsoft/Google sign-in are available in every tier. SAML single sign-on and SCIM are planned.
For teams that want fast, reliable & actionable insights, while keeping data access controlled.
Medium-sized companies in the UAE, then the GCC, that run analytics on databases they maintain and must withhold rows from their own staff or partners.
Evaluations are free and run on sample, synthetic or anonymised data.
