Private AI for the documents that run your business
SATORI reads every document your organization runs on: contracts, specs, policies, tickets, spreadsheets, scans and photos. CARL answers your team's questions from them, with the exact line attached, and never leaves your private appliance.
satori · a Japanese word for sudden understanding: the moment everything clicks and you see the whole picture
First deployment: 7,501 files on a live construction job · each dot is a document · drag across the field
One construction project alone: 7,501 files, more than 70,000 printed pages.
The apps
Every app reads from the same checked record and runs on the same private appliance. Each demo below is loaded with a fictional sample project, so you can click around right now.
The whole project record as a dot cloud. Ask in plain English; every quote opens at its source line.
Try the demo →
The job schedule as a calendar the field actually reads: today, the week, milestones, deliveries.
Try the demo →
Click a one-line drawing and mark each piece of equipment through its inspection stages.
Try the demo →
The job's report card: grades, odds and trends, counted from the job's own files.
Try the demo →
Scan calibrated tools in and out. Who has it, where, and when its calibration runs out.
Try the demo →
PDF markup and takeoff for heavy drawing sets, with no account and no cloud.
Try the demo →
The whole job's roster: who, which crew, hours, and what each person was issued.
Preview the demo →
Lockout/tagout on the same one-line map: isolation points, locks and checks.
Demo coming soon
Details when it ships.
Demo coming soon
Walk the job in 3D: a camera walk of the space becomes a room you can move through and inspect up close, with pins tied to the job, area and sheet.
Demo coming soonEvery app in detail, with what it replaces and what that costs →
Ask CARL
Compute Analysis Reasoning Logic
His own private engine on a dedicated appliance. He reads every page you give him, with no token meter and nothing sent to an AI company.
Takes apart documents, spreadsheets, scans and photos. Counts whole ledgers, not samples, and finds where documents disagree.
Thinks it through the way your best expert would: which document governs, which version is newest, what is still open.
Every answer follows from your documents. Every quote is checked against its file, and when the answer isn't there, he says so.
These are real questions from SATORI's first deployment, a live construction job, with CARL's actual answers, trimmed for length, with owner, contractor, supplier and file names removed. Pick one and watch the record light up.
It says when it isn't there
Every quotation in every answer is checked against its source file before you see it. When the documents don't hold the answer, CARL says so.
"Transformers typically carry a standard 5-year manufacturer warranty."
An illustration of the typical confident guess. It comes from nowhere in your documents.
"The transformer-specific warranty period isn't in what I have read. The only period I can stand behind is the 1-year end-to-end Contractor warranty for the Work."
6 of 6 quotes verified. Nothing invented.
Measured on a live job, not a demo
First deployment: one electrical scope on a hyperscale data-center campus, 9 September to 3 November 2026. Built independently, proven in the field.
120 people × 5.5 hours a week hunting for information × 40% taken over × 48 weeks × $119 an hour: about 12,700 hours handed back.
120 people from the office to the foreman, six 10-hour days. Search time, rework, RFI effort and AI seats, from the value model.
Estimates from published industry figures, not measured savings: 5.5 hours a week searching for project data and 52% of rework traced to bad data (PlanGrid & FMI, 2018), about 5% rework (CII), $1,080 per RFI (Navigant, 2013), and $119 per labor hour on the reference job. The peak figure is the base case of the calculator below, where you can put in your own numbers.
Field figures are every real question asked from 9 September to 3 November 2026, with nothing removed (CARL's own practice runs and test runs are not counted); they measure whether each quotation matches its source file. Answer times are the median of every timed real answer. The graded test is separate: 400 questions written with known answers from the job's documents, 28 to 29 September 2026. First pass 370 of 400; every miss was traced, fixed and re-asked, for 392 of 400.
Private by design
Copilot, ChatGPT, Claude, Gemini and Grok all process your clients' confidential documents on the AI company's servers. Many of those companies are hyperscale data-center owners themselves: the clients this industry builds for, or their direct rivals. SATORI sends them nothing.
No document, passage or question goes to OpenAI, Anthropic, Google, Microsoft, Amazon, Meta or xAI.
The appliance opens no ports to the internet. People reach it only through a private encrypted tunnel, after a work-email sign-in.
When a client's security team asks how your AI handles their documents, you have a clean one. Not legal advice; have counsel read your NDAs.
What it's worth
Worked example from the first deployment, construction. It starts at a team of five on industry medians. Pick a larger preset or move the sliders to your own numbers.
hours a year handed back to the work, full-time people's worth of searching.
A model with cited industry inputs, not a measured result. Your pilot replaces the assumptions with your own numbers.
Nothing else like it
In September 2026 we checked 173 AI products across the US, UK, Europe, Australia and Asia against these five capabilities.
Every reference between documents, linked before you ask.
Every quotation checked against its file. Failures are shown to you.
Newest revision wins. Whole ledgers counted. Conflicts surfaced.
Its own model on its own appliance, off every AI company's servers.
Every miss your team reports becomes a permanent test.
One engine, three sizes
Independent intelligence index, published scores. The same CARL at every size: moving up is an appliance upgrade, not a new product.
In the first deployment now. Beats a flagship cloud model at reading documents, charts and photos.
Faster answers, many at once, for whole project teams.
In the range of the big cloud models, and still private.
It gets better
Every answer is saved with its receipts.
Verified answers join the gold set: 609 so far.
Graded against the documents, root cause found.
A new test for every miss. It can't quietly come back.
Nothing ships if it scores worse.
Go deeper
Partnership
A partner gets a private appliance run and kept sharp by SATORI, and exclusive use of SATORI and CARL in their market and industry while the agreement holds.
While it holds, SATORI is not offered to anyone else in your market.
A private appliance run, upgraded and kept sharp by SATORI. You fund it; SATORI runs it.
Your field's questions decide what gets built next.
Already on a SATORI team?
Use your work email. We'll send a one-time code; no password to remember. Only people on your company's access list can get in.