AI where your people already work, that shows its numbers, only suggests, and stays under your company’s control.
AI that cannot show its work is not used.
An answer without its source is a guess, and engineers will not act on it.
Security committees also ask what is sent, to whom, and whether the AI can be turned off.
BeforeAn AI answer with no source.
With OneStopThe tag, the period and the calculation behind each number.
BeforeHandover notes, work orders and texts typed from scratch.
With OneStopA draft in seconds, edited and approved by a person.
BeforeUnknown data flows to an AI provider.
With OneStopMinimal data sent, a published provider, and a switch per company.
The AI sits in the apps: a panel beside the data, a button on the page.
Only what the answer needs is sent, mostly plant values, tags and events; names and emails are left out where possible.
Answers come with their numbers and sources, and stream with a Stop button.
Each suggestion is accepted, edited or dismissed; the AI never writes data on its own.
Each screen below is from the Demo section of OneStop, on a simulated mine. Ask us for a visit to try them yourself.
The number, its tag, its period in the mine’s time and the PI calculation behind it, with a trend of the values.
Ask Filon answers from the site’s data model and shows how each number was computed.
A control loop’s checks explained in plain words, with "Diagnose with AI" and "Ask PI about this" beside them.
Diagnose an asset, explain a control loop, suggest a downtime reason, draft a work order or a handover note.
The OneStop apps share one sign-in, one subscription and one data model, Filon. Here is what this app exchanges with the others.
Ask PI on your PI data, Ask Filon on the site’s data model, "Ask about this page" in about ten apps.
An incident, a control loop, a stop, an asset’s health, refused rows of a data feed, a failed connection test.
Health briefs, shift weather briefs, plant and fleet summaries, period analyses, handover notes.
Work orders to paste in your maintenance system, support requests, clear text messages, report pages.
A downtime reason, the PI points of an asset or a loop, CSV columns to match, alarm fixes.
Answers about PI and AVEVA cite the AVEVA documentation, with links.
Yes. The AI is set per company: when it is off, nothing about your company is sent to the AI provider, and every app keeps working. New companies start with it off until they accept the provider.
Only what the answer needs, mostly plant values, tags and events. Names and emails are left out where possible, and secrets are removed from what people type.
Only by the features that say so on their page, such as Ask PI and Filon, and it is encrypted with your company’s own key. Usage logs keep no conversation text.
No. It only suggests; a person accepts, edits or dismisses each suggestion.
What it reads, what it writes and who sees it. The whole picture: Security and trust.
New companies start with the AI off until they accept the provider.
Only what the answer needs is sent.
Kept conversations are encrypted with your company’s own key.
What is true of this app is true of every app of TTech OneStop: your data, your people and your language come first.
The apps read your data through your own AVEVA CONNECT and write back only their own results, in their own streams. Your existing data is never changed.
OneStop runs in the browser. Your administrator sets up one connection to your AVEVA CONNECT, once, and every app uses it.
Your people sign in with their Microsoft work account. Your own admins decide who uses which app.
Every score, alert and AI answer shows its reason and where its numbers come from, in plain words.
Every screen, every message and every report, in the language each person chooses.
Subscribe to the apps you need and add others later. Try each one first in the demo, on ten simulated mines.
Book a demo on our simulated mines, or tell us how you want to use it on yours.