The local AI that prepares your questions before the analyst starts.

You run a team, or the whole company. The numbers are yours to answer for, but analysis is not your day job. You know there is something in the margins, and “can you look into it?” is not a brief, because every round trip with the data team or an outside analyst costs days. Take the first look yourself. Open your own exports in Totsum, ask in plain language, see what the data actually holds, and shortlist what deserves the real work.

What you hand over is a follow-up report with your questions, the charts and the queries, built to be verified by the person you are briefing. The real analysis starts with the right questions, and it starts days earlier.

Totsum requires a GPU with roughly 12 GB of VRAM, or a Mac with an M3 Pro chip or better and at least 18 GB of unified memory. Please check compatibility before you apply.


How it works

Ingest. Explore. Validate.

Everything in Totsum runs through one loop. Your files go in, you put your questions to them, and a reviewed report comes out.

FIG. 1 The loop runs through the three panes of the app. You ingest a file, you explore it with questions, and you validate the keepers through a report.

“Ingest” is the data world’s word for bringing a file into a tool. You drop a file in, and Totsum reads it into tables that you can question. The file itself stays where it is, unchanged.

It explores your data, question by question.

The first hours with a new file are the expensive ones. You don’t know yet where to look, and the usual tools charge you for every hunch, because each question means building an export, a pivot table, a formula or a chart. In Totsum you ask, you look, and you ask again. A hunch costs you minutes rather than an afternoon.

The kind of questions

  • “Which ten products drive most of the returns?”
  • “How did weekend sales compare to weekdays, month by month?”
  • “Which country’s customers spent the most?”Even when the customers and the invoices live in two different files, Totsum looks for the relevant tables and joins them.

The kind of answers

Every answer comes back in a readable form. It brings a plain-language answer, the chart, a preview of the rows behind it, and the query that led to them. When Totsum notices that it had to make an assumption in order to answer, such as which file, which column or which spelling to use, it says so with the answer, and you can correct it. You look at the result, then you shortlist it or rule it out, and you ask the next question.

The AI Assistant panel in Totsum: the question "Which ten products drive most of the returns?" answered in plain language with a link to the chart, and a "Try asking" list of further questions.

Except when it is too complex, the generated SQL is also translated into a query flow, which gives a step-by-step account in plain language of what was read, filtered, grouped and calculated.

The AI runs on your machine.

The files that need AI the most are exactly the ones you can’t hand to a cloud AI. Client data under NDA, personal data and anything a contract says must stay home all fall into that category. With Totsum the AI comes to the data instead. The language model and the SQL engine run on your own computer, and your file is opened where it already sits. That’s also why Totsum asks for a serious machine. The AI lives with your data, rather than the other way around.

See it work

Watch it work with the connection off.

The demo shows the whole point in one take. The network goes off on screen, and then a confidential file gets its first look with the connection switched off. You see the answer, the chart and the rows behind it appear.

Watch the network go off first, and then the preview appear. Totsum can work on a computer with no network connection.

The rhythm

It works while you work.

A real AI model doing real work on your machine takes real time. Expect a few minutes for an answer rather than a few seconds. Totsum is built for that rhythm rather than against it. With the usual tools, you stay at the keyboard for as long as the digging takes. With Totsum, you ask the question, you go back to whatever else you were doing, and the answer is waiting when you come back, with the chart, the preview and the rows behind it. The demo above shows the wait instead of hiding it.

Privacy

Your data is processed on your machine.

Totsum processes your datasets, questions, generated queries and results on your computer. Where the confidentiality of your work requires it, Totsum can work on a computer with no network connection.

THE INTERNETCAN STAY OFFYOUR COMPUTERYour fileCSV · Excel · JSONLOCAL AI MODELWrites the queryLOCAL SQL ENGINERuns it on your fileRESULTAnswer · chart · query flow
FIG. 2 Your file, the AI model and the SQL engine run inside your computer. There is no account, no sync, no telemetry and no update check.
  • Local AI model. The language model runs in memory on your machine rather than on a cloud API.
  • Local SQL engine. Every query runs in an embedded database engine on your computer.
  • No account, no sync, no telemetry, no update check. There is nothing to sign up for, and your license is verified on your device.
  • Can work on a computer with no network connection. The installer downloads the AI model once (~7 GB). On a machine with no internet access, the model files can be provisioned by hand instead, and the machine can stay disconnected.

The handoff

A report built for the human check.

AI is genuinely helpful, and that’s exactly the trap. After a few good answers in a row from any AI, it’s human to let your guard down and stop checking. But sooner or later an answer is wrong. That is a certainty rather than a small risk, and when a number is wrong, the decision belongs to a person, never the AI. So Totsum doesn’t deliver a final report. It delivers a follow-up report, addressed to the expert or the team who will do that crucial last step. The report carries what they need in order to check the work. It holds the question exactly as you asked it, the chart, a subset of the data behind it, the query flow and the generated SQL. It also carries a verdict field for each result, and the verdicts come back with your reviewer's name on them.

Page one of a follow-up report with a company letterhead: the report basis, the reviewer’s name, the original question, a note for the reviewer, a Valid verdict ticked, and the revenue-by-country chart.

It says what it is

The report opens by describing its own standing, in its own words: “These results come from exploring source data with AI assistance. They are a first look rather than a validated analysis: each one is recorded here for a qualified reviewer to judge, and the verdicts against them are that reviewer's.”

Your reviewer fills it in

The report exports with fields to complete. Against each result your reviewer records Valid, Invalid or Inconclusive. The verdict is theirs, and it carries their name rather than ours.

The reviewer is the person you choose to check the numbers before anyone relies on them: a colleague, your accountant, an analyst, or you, if that is your competence. The reviewer is not an AI, and not someone Totsum supplies.

The verdicts come back

When you import the filled review, the verdicts land back in the app beside the results they judge. You keep a record of what was checked, by whom, and what they concluded.

Totsum never marks a result Valid. We built it so that the judgment is a person's and is recorded with their name, because a number without a name behind it is a liability. An Invalid verdict is not wasted work either. Understanding why an answer was wrong usually teaches you something real about the data, such as what a column actually holds, or which assumptions the dataset makes without saying so. What Totsum can't do is written down too.

One-time purchase

Pay once. It’s yours.

Most data tools today are metered. They charge per seat, per month, or per question you dare ask your own files. Totsum is a one-time license. Because the AI runs on your hardware, there is no per-query cloud bill to pass on. No subscription also means that no meter runs behind it. There is nothing to cancel, and the app you bought keeps working on your machine.

  • The license is perpetual. It costs €249 for Personal or €499 for Commercial, and it is sold in the European Union (your currency and tax are set at checkout).
  • Any updates we release for your major version come free, with no extra fees.
  • There are no monthly charges, no seat renewals and no surprise price increases.

Compare plans

Before you join the beta

Totsum runs a real AI model on your machine, and that is why the hardware requirement is real. The model needs a lot of memory, and that memory has to be on your computer, because the model runs there, next to your files. Most current computers do not meet the requirement. See the hardware compatibility guide to check yours.

MacApple Silicon, M3 Pro or better18 GB unified
WindowsDedicated GPU — NVIDIA, AMD or Intel12 GB VRAM
AI modelDownloaded during installation, runs on your machine~7 GB on disk
NetworkCan work on a computer with no network connection
AI outputA first look; verdicts are recorded by your reviewer, never by the AI

A private beta opens soon. A beta is the step before an app goes on sale. A small group of early users runs it first, free of charge, and what they run into shapes the finished product. We call it private because you join by applying rather than by downloading.

If your machine qualifies, apply with your email. Places are limited, so not every applicant can be contacted. Beyond the confirmation e-mail, you’ll hear from us only if you’re selected.

Apply for the private beta