Alpha DecisionHub
Upload a spreadsheet. Get an answer.
DecisionHub turns a CSV or Excel export into a working dashboard in seconds, then lets you ask it questions in plain English. No formulas, no modelling, no analyst — and every number on screen computed from your own rows.
£1.24M
4,812
£258
Revenue by month
Ask your data
AIJune revenue rose 33% on May. Two accounts — Northfield and Carrow — placed first orders above £40k, together accounting for £81k of the £94k increase.
The problem
Most businesses are data-rich and answer-poor.
The spreadsheet holding the answer is usually already on someone's desktop. What's missing is the hour, the skill and the person to interrogate it.
The data is there. The answer isn't.
Every business runs on a handful of spreadsheets that nobody has time to properly interrogate. The numbers exist; the insight is still locked inside them.
Pivot tables are a skill, not a given
Getting from a raw export to a chart worth showing someone takes an afternoon and a person who knows their way around a formula bar.
Analysts are expensive and busy
Small teams can't keep one on staff, and the ones who exist are booked. So the question goes unanswered, and the decision gets made on instinct.
How it works
From a file on your desktop to a decision.
Four steps, none of which involve a formula bar.
Upload your file
Drag in a CSV or Excel export. Column types are detected automatically — what's a number, what's a date, what's a category.
A dashboard appears
KPI cards and charts generate themselves from what's actually in the file. No configuration step, no blank canvas to fill in.
Ask in plain English
"What are the top five regions by revenue?" Type it as you'd say it and a chart comes back, with the query it ran shown underneath.
Ask why, not just what
Follow up in a conversation that remembers the thread, and get a written explanation with charts dropped inline where they help.
What you get back
A dashboard, and somewhere to ask about it.
The KPIs and charts build themselves from what's in your file. The conversation sits next to them — ask a follow-up and the answer cites the rows and columns it came from.
- KPI cards and charts generated from your own columns
- Plain-English questions, with the query shown underneath
- Every figure computed from real rows, never invented
£1.24M
4,812
£258
Revenue by month
Ask your data
AIJune revenue rose 33% on May. Two accounts — Northfield and Carrow — placed first orders above £40k, together accounting for £81k of the £94k increase.
Two ways to build a dashboard
Use AI when it earns its place — not before.
Statistical mode is pure computation and needs no AI at all. AI mode adds judgement about what matters and writes the summary. Both draw their numbers from the same rows.
Statistical mode
Pure computation — no AI call, no waiting, nothing leaves the page. Totals, column types and charts are worked out directly from your rows and re-render instantly when you adjust them.
AI-generated mode
The AI decides which measures actually matter for your data, picks the chart types, and writes an executive summary with specific insights. The figures underneath are still computed from real rows.
What you get
A working analyst's output, without the analyst.
CSV and Excel, straight in
Drag-and-drop upload for CSV, XLSX and XLS — up to 50 MB and 5,000 rows per file today. Keep several data sources side by side, preview the raw rows before you commit, and switch between them instantly.
Charts chosen for you
Line for a trend over dates, bar for a breakdown by category, pie for a split, area for several measures at once — picked from the shape of your data, not from a menu.
A conversation, not a query box
Multi-turn chat with history kept per data source. Answers stream in as they're written, with charts embedded mid-explanation and starter questions drawn from your own column names.
Guarded against invention
Any column the AI refers to is checked against the ones that actually exist in your file. The narrative comes from the model; the numbers come from your rows.
Nothing is thrown away
Uploads, generated dashboards and chat history are stored against your account, so you can come back to a dataset and pick up exactly where you left it.
Take it into the meeting
Export any chart as an image, and see the query behind every AI answer — so what you present is something you can also explain.
AI you can take into a board meeting
The figures come from your rows. Always.
An analytics tool that invents a plausible number is worse than no analytics tool at all. So DecisionHub draws a hard line: the AI chooses what to show you and explains what it means, but every value on the screen is calculated from the file you uploaded.
Computed, never guessed
KPI totals and chart values are worked out from your actual rows — the model never supplies a figure.
Checked against your columns
If the AI refers to a field that isn't in your file, it's caught before it ever reaches the screen.
Show your working
Every AI answer displays the query behind it, so you can sanity-check the result rather than take it on faith.
AI is optional
Statistical mode produces a full dashboard with no AI call at all, and the model itself is switchable — including options that run locally.
Who it's for
Built for the businesses BI tools skip.
Small and growing businesses
Running on exports from a till system, a shop platform or an accounting package, with nobody whose actual job is analysis.
Operations and finance leads
Who already know their numbers intimately but lose hours rebuilding the same monthly view by hand, every month.
Teams without a data team
Where the question is urgent, the spreadsheet is right there, and waiting a week for someone technical isn't an option.
The same principle runs through every Alpha product: AI does the volume work, and a qualified human keeps the final say.
Questions
Answers before you ask.
No. You upload a file and a dashboard appears. When you want something specific, you type the question the way you'd say it out loud. DecisionHub shows you the query it ran underneath the answer — useful if you're technical, ignorable if you're not.
No, and that distinction matters. Totals and chart values are computed directly from the rows in your file. The AI decides which measures are worth showing and writes the explanation around them — and any column it refers to is validated against the columns that genuinely exist in your data.
Yes. Statistical mode runs entirely on computation — dashboards, KPI cards and charts with no AI call at all. The AI features are opt-in, and the underlying model is switchable, including options that run locally.
CSV, XLSX and XLS, up to 50 MB, with up to 5,000 rows parsed per file. Larger files, live database connections and direct integrations with tools like accounting or e-commerce platforms are on the roadmap.
Traditional business intelligence tools assume someone will model the data and build the dashboard. DecisionHub assumes nobody will. It's aimed at the point before BI makes sense — where the data is a spreadsheet and the alternative is not asking the question at all.
It's the furthest along of our products — the full workflow is built and running. We're onboarding early users now. Book a demo and we'll set you up with your own data.
See DecisionHub on your own spreadsheet.
Book a 20-minute demo. Bring a real export and we'll build the dashboard live.