- contact@insightaiviz.com
InsightViz · AI-native business intelligence
When a number moves, ask the dashboard why
InsightViz is business intelligence where the dashboard explains itself. When a figure moves, press Why: it works out what drove the change — which region, which product, which segment — shows the evidence, and answers the follow-up questions in plain words. Every figure it writes is traced to your rows, every answer is bounded by your definitions and access rules, and the query behind it is one click away. Your Tableau workbooks come across; your data stays in your own cloud.
- Explains any chart in one click
- Grounded in your rows, working shown
- Runs in your cloud; imports Tableau
Runs in your cloud · your data never leaves it · your sign-in
Ask why
The question every dashboard leaves unanswered
A chart shows that sales fell. It never says why. InsightViz works it out from the rows — no guessing — and lets you keep asking.
Explain this
Press Why on any chart or headline number. InsightViz picks the period that moved and the one to compare it with, breaks the change down by every business dimension, ranks what drove it, and writes it down in plain sentences — each figure traced to a cell.
Ask about it
Ask the follow-up in words: was it volume or price, is this unusual, what moved the other way. The answer uses only the evidence on the panel plus at most one further governed query — and says so when the rows cannot answer.
What changed
Every panel’s latest period against the one before, ranked by how much it moved — on the dashboard, on your Home page and in the scheduled email. The next step is offered with it: filter to the driver, break it down further, set an alert, or keep the question as an insight.
Insights
From one answer to the story the team reads every week
Why? explains one chart, once. An insight keeps the question: several parts, each a governed query, written up and checked figure by figure — then re-run for whoever opens it, at the filters they have on.
Make it an insight
Under every Why? explanation: keep this question. It becomes a story of its own — the trend, the breakdowns that drove it and a written synthesis with what to do — that re-runs whenever the data does.
This week
One standing story per dashboard: every panel’s latest period against the one before, the parts that moved most, and what to do about them — ready before the Monday meeting.
Proposed, or from a brief
Let the AI propose questions worth asking of a dashboard, or write a brief in your own words. You see the plan before anything runs, and nothing is published until you say so.
Checked before it is published
Every figure in the story is matched to a row and marked. Click one to see where it came from, then open it further by region, by category or over time. A story with untraced figures is refused at publish — or published with that recorded on the version.
On the dashboard
A published insight sits on the grid like any panel and follows the page’s filters: pick East and the story is about East. A slice already worked out comes back from cache.
Watch it
Alerts mail you when a number crosses a line — one is proposed by every Why?. Subscriptions send the page on a schedule, with What changed and the published weekly story in the body of the email.
What it does not do. A story is only as good as the share of its figures that trace, so that share is printed under every one. It explains what the rows show; why a customer ordered less is still a question for a person.
Why InsightViz
Three things the old BI stack never got right
Dashboards became files nobody could review. The same metric meant different things in every workbook. And asking a new question meant filing a ticket. InsightViz fixes how the business is described, not just how a chart looks.
Ask
An assistant that only speaks your business language
Ask in plain words. The assistant works only from your approved business definitions — the metrics and dimensions your teams agreed on — so an answer is always the company’s number, never a guess.
- Plain-language questions, real numbers
- Every answer shows how it was worked out
- People only ever see what they are allowed to
- Keep an answer as a chart on any dashboard
Build
Dashboards that are easy to build, review and trust
Every dashboard is a readable document, not a locked file. See exactly what changed between versions, restore any earlier one, let the assistant draft a new one, and move it from draft to published with a note.
- Version every publish; restore any
- Build from Explore or from an answer
- Filters that use your business dimensions
- Every chart explains how it was computed
Connect
Your data, in your cloud, always fast
Connect databases, warehouses and files. InsightViz keeps a governed copy in your own cloud storage, so dashboards load instantly and your source systems are never slowed down by reporting.
- Databases, warehouses, CSV and spreadsheets
- Scheduled refreshes; roll back to any earlier one
- No load on operational systems
- Mark sensitive data as never-leaves-the-server
How it works
From your data to a question answered
Five steps, each one visible and auditable.
1
Connect
Add a database, a warehouse or a file. Credentials are stored securely and never leave the server.
2
Refresh
A governed copy of the data lands in your cloud. Every refresh is a new version; the previous one is kept for rollback.
3
Define
A business model is proposed automatically: the dimensions and metrics, with business names. The assistant fills in descriptions.
4
Explore & build
Pick fields, see the chart and the table, keep it, add it to a dashboard. Or ask the assistant to do it.
5
Govern
Workspace roles, row and column access rules, and a full record of who asked what — including what was refused.
Building dashboards
Four ways to get a dashboard — and they all end in the same place
A dashboard is a document here, so it does not matter who wrote it. Describe what you want, show a picture of it, place the panels yourself, or bring the workbook you already have. All four produce the same reviewable, versioned document, using the same business definitions.
Describe it
Say what you want in a sentence
“A monthly sales overview for the leadership meeting — headline numbers, the trend, and a breakdown by region and category.” You get a laid-out dashboard back. The assistant may only use fields your model actually has, so it cannot invent a metric; if it wanted one that is missing, it says so rather than shipping a panel that fails.
- A dashboard from a plain-English brief
- Only ever uses your approved fields
- Nothing is saved until you press save
- Edit any panel afterwards, by hand
From a picture
Show it what you want
Drop a screenshot of a dashboard you already have — Power BI, Tableau, a slide, a whiteboard photo. It is read panel by panel and rebuilt over your model. Anything the picture asks for that your data does not have is listed beside the preview, never quietly swapped for something else.
- PNG or JPEG from any tool
- Rebuilt over your own fields
- What it could not match, listed
- Edit it like any other dashboard
Design it
Drag the fields, keep the control
The same field picker your analysts use in Explore: dimensions, metrics, time grains, filters, chart type — with the data previewing as you build. For the people who know exactly what they want and would rather not describe it.
- Point-and-click, with a live preview
- Raw fields or ones you derive
- Joins and filters you define once
- The panel’s definition is visible, always
Import it
Bring the work you already did
Drop a Tableau workbook. The data, the field names, the calculations, the sheets, the layout, the filters and the click-to-filter behaviour all come across — with a report of anything that needs a second look.
- Whole workbooks, not just data
- A fidelity report, sheet by sheet
- Nothing dropped quietly
- Repoint it at live data afterwards
See it in action
The console your teams actually work in
The same workspace for the analyst building a model, the manager asking a question, and the executive reading the result.
For the people who only read the dashboard
Self-service that does not mean “go and learn a tool”
Most of your licence spend is people who open a dashboard, filter it, and want the number. That should not require training, a ticket, or a seat.
Filter and drill without asking anyone
Filter controls come from the business dimensions, so they read like the business. Click a bar and the rest of the page follows it. Click again to go a level deeper — region to category to product — and the page works out which level that is.
Take it into the meeting
Any dashboard downloads as a PDF or an image, sized to the dashboard rather than chopped onto A4, so nothing lands half-cut across a page break.
Ask the dashboard a question
The assistant is on every page. Ask a follow-up the dashboard does not answer, get the number and a chart, and keep it as a new panel if it turns out to matter.
See how any number was worked out
Every panel shows the question it asked, the query it ran, and which access rules applied. When someone challenges a figure in a meeting, the answer is on screen rather than in somebody’s head.
Tableau migration
Drop the workbook. Keep the work.
Years of decisions live inside your workbooks — field names, calculations, filters, layouts. InsightViz carries them across and tells you exactly what needs a second look. Power BI files are on the roadmap; a screenshot of a Power BI page can already be rebuilt from a picture.
Data comes across
Every table in the workbook’s extract is brought into your own cloud storage, catalogued and ready to query at once.
Definitions come across
Field names, roles, aggregations, hierarchies and calculated fields become your business model. Calculations convert automatically — including table calculations and fixed level-of-detail expressions. The handful that cannot be settled without a specific view are flagged for a person.
Dashboards come across
Worksheets become charts and dashboards keep their layout, so the page your team opens is the page they know.
Filters come across
The filters on each sheet, the controls on the dashboard and the click-to-filter interactions all carry — so the numbers match and the dashboard still behaves the way people expect. Anything that cannot be carried is named in the report, never dropped quietly.
| Tableau / Power BI today | InsightViz | |
|---|---|---|
| Dashboards | Locked files; nobody can see what changed | Readable, versioned documents the assistant can draft and edit |
| Metric definitions | Per workbook; drift across teams | One set of business definitions, used by every dashboard and every AI answer |
| Ask a new question | Rebuild a sheet, or file a ticket | Ask in plain words; keep the answer on a dashboard |
| Why a number moved | An analyst’s afternoon, then a slide | One click on the chart; keep it as a story that re-runs every week, every figure traced |
| Migrating what you have | Rebuild by hand, sheet by sheet | Drop a Tableau workbook: data, definitions, dashboards and filters, with a report of what to review. Rebuild any other page from a screenshot |
| Who sees what | Set per workbook, easy to forget | Set once on the data; applied to every dashboard, exploration and AI answer |
| Cost to view | Per-seat licences for people who only look | No viewer tax — the data is in your cloud and the engine is included |
| Where the data lives | A vendor’s format on a vendor’s server | Your own cloud storage, in an open format any tool can read |
| Proving a number | Open the workbook and read the calculation | Every panel shows the question, the query and the access rules that applied |
Security & governance
Built for the questions your security team will ask
Identity, access and evidence — designed in, not bolted on.
Your single sign-on
Okta, Entra, GoogleRoles from your directory
Row & column access
Set once per datasetApplied everywhere
Full audit trail
Who asked what, whenRefusals recorded
Your cloud, your keys
AWS · Azure · GCPData never leaves
From the blog
Latest from our blog
Notes on trusted metrics, dashboards you can review, and what migrating off Tableau really involves.
AI
The Report That Arrives Before The Meeting
A dashboard waits to be opened. What it takes for the numbers, the reason they moved and what to do about it to arrive first — and why the email only carries what was checked.
Read article
Migration
Rebuild A Dashboard From A Picture
A screenshot of the dashboard you have is a better specification than any brief. What gets read, what gets rebuilt over your own data, and what it refuses to pretend.
Read article
AI
Why Did Sales Drop? The Dashboard That Explains Itself
Every chart can show that a number fell. None says why. What it looks like when the dashboard answers in one click — worked out from the rows, not guessed.
Read article