Business intelligence that learns your business.
Connect the systems you already run and ask in plain language. Mercury11 answers, builds the dashboards and writes the reports, all on your company’s own rules. No warehouse project, no data model to maintain, no per-seat licenses, and no AI deciding on its own what your numbers mean.
No credit card. The trial starts at your first answer, not the day you sign up.
Ask Metis
Three customers made 61% of gross margin last quarter. One of them is slowing down.
Cedar Ridge is down because their hardware orders stopped in August; service revenue is flat. Under the net 90 rule their July invoices count in Q3, so the drop is real, not a timing effect.
Save to a dashboard
Findings how Metis is reading your data
Metis states the assumptions it is making so you can correct them. Only real conflicts between systems become questions.
Ask. Save the answer as a tile. Correct how Metis reads your data in one click. No model, no SQL, no ticket.
Three ways companies buy business intelligence today
All three can draw a chart. None of them knows that hardware is on net 90, that half of July was a re-import, or that marketing and finance count a customer differently. That knowledge is what makes a number right, and it is the part nobody sells.
A warehouse, a data model, a consultant. The business rules go into a semantic layer the people who run the business cannot read or change.
Time to first answer
Months
Power BI, Tableau, Looker. Every person who wants to ask needs a seat, so most people never ask. They ask the one person who has one.
Who can ask
Whoever has a seat
Connect your data and ask. Fast, and it decides for itself what counts as revenue, spend or a customer. Two reasonable prompts get two reasonable answers.
Who defines the metrics
The model
The hard part of BI was never the chart. It was agreeing on what the numbers mean.
Time to first answer
Minutes
Rules instead of a data model. Everyone asks. The company defines the metrics.
Connect the systems and Mercury11 starts answering on what it finds, stating the assumptions it makes so the people who know can correct them in a click. Those corrections become rules, written in plain language with a name and a date, and every answer, dashboard and report runs on them. No model to maintain, no seat to buy, no AI guessing.
Business intelligence that knows your business, because your people taught it.
How it works
Connect what you already use
Sign in to your tools or paste a read-only key. Books, billing, CRM, ad platforms, analytics, field service. Drop an export for anything without a connection. A warehouse, if you have one, is one more source.
Metis learns how your business works
It reads the data, states how it is interpreting it, and finds where your systems disagree. You confirm or correct in plain language. Each answer becomes a rule with your name on it, and the rules are the data model: readable, editable, and applied everywhere.
Ask, build, schedule
Anyone asks in their own words. Any answer becomes a dashboard tile. Any question becomes a scheduled report. All of it shows the sources it read and the rules it applied, so the number can be checked by anyone.
Works with the tools you already pay for
- QuickBooks
- Xero
- Stripe
- Square
- Google Ads
- Meta Ads
- Google Analytics
- Search Console
- HubSpot
- Salesforce
- Shopify
- ServiceTitan
- Jobber
- CallRail
- Google Sheets
- Snowflake, BigQuery, Databricks (Enterprise)
- Anything else: drop an export and it counts
A semantic layer your operators can read.
Traditional BI puts the business rules in a model only the data team can touch. AI tools skip the rules entirely. Mercury11 writes them down in plain language, by the people who know, and applies them everywhere.
Viewers ask. Editors confirm how Metis reads the data and write rules. Admins settle conflicts and set company-wide definitions. When two rules disagree, nothing applies until someone with the authority decides. Then every answer, dashboard and report uses the same definition, no matter who is asking or how they phrase it.
Every output lists the sources it read and the rules it applied. If a rule changes, the next report says which one and what moved because of it.
Everything you bought BI for, on one set of rules
Plain-language answers
Any question, any team, in their own words. Sources and rules shown under every answer.
Describe it, or save an answer
"Build me a leadership dashboard." Or save any answer as a tile. Refreshes on every sync; no tile ever disagrees with an answer.
Scheduled, with the why
Weekly or monthly, by email with a PDF, written from the data with what changed and what is worth asking.
What changed and what disagrees
Shifts, duplicates, system tie-outs and the assumptions Metis is making, ranked by how much money they touch.
Mercury11 next to the BI you know
| Traditional BI | AI chat tools | Mercury11 | |
|---|---|---|---|
| Setup | Warehouse, data model, consultant. Months. | Upload an export and ask. Minutes, but you do it again every time. | Connect the systems once. Your data updates on its own from then on. |
| Keeping it current | Pipelines somebody maintains. Breaks when a column moves. | Not plumbed in. It knows whatever you dropped in, as of when you dropped it, and forgets when the chat ends. | Connected and synced on a schedule, every source unified in one place. Every answer says what time the data is as of. |
| Who defines the metrics | The data team, in a semantic layer the business cannot read. | The model, on its own, differently each time. | Your people, in plain-language rules with their names on them. |
| When systems disagree | Reconciled by hand in the model, or not at all. | Averaged, guessed, or not noticed. | Found, explained, and settled by the person who knows. |
| Who can ask | Whoever has a seat. | Whoever has a login. | Everyone in the company. One price, not per seat. |
| When a definition changes | A ticket to the data team. | Nothing; it never knew the old one. | Edit the rule. Every answer, dashboard and report updates, and the next report says so. |
| Can it explain why | No. Someone in the room has to. | Sometimes, from the data alone. | Yes, from the data and from what your people have taught it. |
Questions people ask on the first day
Connect three systems. Ask the question you have been wanting to ask.
No warehouse, no model, no consultant. Mercury11 answers on what it finds, tells you how it is reading your data, and gets better with every correction. Free to start; the trial clock runs from your first answer.
Questions people ask first
Is this a replacement for Power BI or Tableau?
For most companies under a few hundred people, yes: the dashboards, the reports and the questions all live here, with none of the data model to maintain. If you have a BI team that builds exploratory analysis in Tableau, keep it and connect your warehouse to Mercury11 as a source; Mercury11 becomes the place the rest of the company asks, and the place the rules live.
Do I need a data warehouse?
No. Mercury11 reads from the systems directly and keeps its own copy per business. If you already have Snowflake, BigQuery or Databricks, the Enterprise plan connects it as a source alongside everything else.
Who maintains the data model?
There isn't one in the usual sense. The business rules your people confirm or write are the model: plain-language sentences with a name and a date, listed on one page, editable by the people with the authority to edit them. When a rule changes, every output updates.
How is this different from asking ChatGPT about an export?
Two things. A general AI is not connected to anything: it knows the export you uploaded, as of the moment you uploaded it, and forgets when the chat ends. Mercury11 is plumbed into your systems, syncs on a schedule, and keeps every source current in one place. And a general AI decides for itself what revenue, spend or a customer means, differently depending on how you ask. Metis works inside your company's rules, tells you which rules and sources it used, states the assumptions it is making so you can correct them, and never answers with a number it did not query.
Can people who are not analysts use it?
That is the point. If you can describe what you want to know, you can ask it. Owners, office managers, bookkeepers and department heads are the usual first users; the data team, where there is one, usually ends up owning the rules.
How do you keep answers consistent across teams?
Everyone asks against the same rules. Marketing and finance can phrase a question completely differently and get the same CAC, because the company decided once what CAC means. Conflicts between rules go to an admin; nothing applies until it is settled.
Who owns the data?
You do. Connections are read-only, each business sits in its own isolated database, you can delete any source's data or export every rule and answer in one click, and we never train models on your data.
What does it cost?
One price for the whole company, not per seat. Plans start at $75 a month, month to month, no contract, with a 14-day free trial that begins at your first answer. See pricing.