AI assistant (Cx Assistant chat)

What is the Cx Assistant

The Cx Assistant is the built-in AI assistant. You chat with it in plain language and it can answer questions about your project, search and read your documents, look up assets, issues, RFIs and planner data, and — with your approval — make changes for you, like creating an issue.

It is available to every signed-in user, with or without an active project, and it never has more access than you do: everything it reads or changes is checked against your own project permissions.

Full chat page vs the chat drawer

There are two ways to chat:

• The full chat page: open Assistant from the sidebar (or go to the chat page directly). It has a history sidebar with all your conversations.
• The drawer: on every other page, click the round ✕ button at the right of the header — it expands to read "Ask ✕ Cx Assistant" when you point at it. A side drawer opens next to your current page — it is resizable and pushes the page content aside, so you can chat while you work. The button briefly expands to show its label once per session to remind you it exists. Buttons at the top of the drawer open your chat history, start a new conversation, or open the same chat as a full page.

The drawer is hidden while you are on the full chat page. Both use the same conversations.

How to choose which assistant answers

Cx Assistant has several assistants, each focused on one kind of work. You pick one when you start a chat.

1. Start a new chat (the new-chat button in the history list or in the drawer).
2. In the toolbar under the message box, next to the + button, click the button that shows the current assistant's name.
3. The "Choose assistant" menu opens. Each row shows the assistant's name, a one-line description, and a "Default" or "Beta" chip where it applies.
4. Click the ⓘ icon on a row ("About this assistant") to open its card: "What it can do", "What it can't do", "Try asking" examples, and "Available on".
5. Click a row to select it, then send your first message.

The assistant is fixed for the whole conversation. The picker only shows on a new, empty chat — to switch assistants, start a new chat. If you see no picker, only one assistant is available to you there. A warning icon next to a name means that assistant can make changes that cannot be undone, such as removing members or locations.

The assistants you can choose from

Inside a project, the "Choose assistant" menu lists:

• Commissioning Assistant (Default) — works with this project's issues, assets, documents and scripts, draws and edits 3D layouts, builds dashboards, reads CAD drawings and administers the project (sets and statuses, members, locations, workflows and settings) — each of those as far as your role allows.
• General Assistant — engineering knowledge, web search and your own files; no project workflows.

In My Workspace (outside any project) the menu lists:

• My Workspace Assistant (Default) — your personal files, engineering knowledge and web search; no project data.
• General Assistant.

Every assistant works with your own permissions: the Commissioning Assistant only draws layouts, builds dashboards or administers the project for people whose role grants it (the layouts, dashboards and project administration permissions, granted to project admins by default), and every change it makes is checked like a change you would make yourself.

How to ask a question

1. Open the chat (full page or the "Ask ✕ Cx Assistant" pill in the header).
2. Type your question in the composer and send it.
3. The answer streams in live. While the assistant is thinking or using tools, a thoughts timeline shows what it is doing step by step.
4. To interrupt, click the stop button in the composer (it replaces the send button while a response is running).

Be specific: name the asset, document, or issue you mean, or attach the file directly. A token-usage indicator near the composer shows how large the conversation is getting.

Attach files and documents to a chat

You can give the assistant files to read:

• Click the + button in the composer to pick files from your computer.
• Drag & drop files straight onto the composer, or paste a copied image or file into the message box.
• To use something already in the project, open Documentation and drag a document, a folder or a script from the file tree into the chat drawer — nothing is uploaded again.
• When you are viewing a document or script, the chat may suggest "Attach the document you're viewing" with one click.

Accepted formats: images (JPG, PNG, GIF, WEBP, HEIC), videos (MP4, WEBM, MOV), PDFs, and Office or text documents (DOC, DOCX, XLS, XLSX, PPT, PPTX, TXT, CSV).

Limits: up to 10 files per message, 60 MB per file. Videos can be up to 200 MB, one video per message — send a second video in a separate message. If a file is refused, a message tells you why.

Each attachment shows its upload progress, a remove button, and a retry button if the upload failed. Wait for uploads to finish before sending. Attachments show inline previews and can be opened in a viewer. Files you upload in chat become real documents — find them later in Documentation under the Chat files scope; they are private to you.

Citations and sources in answers

When the assistant bases an answer on your project data, it shows its sources:

• Citations deep-link into documents and scripts — clicking one opens the document at the exact page, or the script at the right section.
• Source chips under a message show which assets, issues, users, or documents the answer drew on; clicking a chip opens the corresponding dialog.

Use the links to verify anything important — the answer should never be more than one click from its evidence.

Tool runs and the effects card

When the assistant works with your project data it uses tools — searching documents, listing assets, creating an issue, and so on. Each tool run appears as a row in the conversation so you can see what it did.

When a turn changes something in the project, an effects card summarises exactly what changed. That is your at-a-glance audit of the assistant's actions for that turn.

The Tasks list: follow a long job step by step

For a job with several steps, the assistant writes itself a task list and shows it to you. A card titled "Tasks" appears just above the message box and stays there while the conversation scrolls.

• The header shows progress as done/total, for example 2/5, with a thin progress bar.
• A pulsing dot marks the task the assistant is working on right now (In progress).
• An empty circle is a task still to do (Pending).
• A green tick is a finished task (Completed).

Click the arrow in the header to collapse or expand the list. The card disappears by itself once every task is completed. The list is the assistant's own plan — you do not edit it; to change course, send a new message.

Research agents: helpers that read in the background

When a question needs a lot of reading — many documents, long reports, several topics at once — the assistant can hand parts of that reading to research agents and write its answer from their reports. A research agent only reads; it never changes anything.

The agents work in the background: the assistant keeps working on the rest of your request while they read, and it can start several for one message. If it has nothing left to do but wait, it may send a short first message and then continue the answer once the reports are in.

Each agent is one compact row in the conversation:

• The row reads "Agent (Researcher)" and a short description of the job.
• While the agent works in the background, its icon pulses and the row says "background task".
• When the assistant is waiting for the agent, the row says "running" with the time so far.
• When the agent ends, the row shows the steps it took, the time and the tokens it used, for example "5 steps · 41s · 2.4k tokens". A warning icon with Stopped, Timed out, Ran out of steps or Failed means it could not finish.

Click a row to open its card underneath, with the steps the agent took. On the card, click "Open thread" to open the "Research thread" window: the outcome, the Briefing the assistant gave, the Steps taken and the final Report.

Limits: the assistant can start up to 9 research agents for one message, and three of them work at a time — the others wait their turn. Once the limit is reached, the assistant does the rest of the reading itself. Stopping the answer, or sending a new message while the assistant is still working, also stops the agents that are still running.

You do not need to do anything — the rows are there so you can see where the answer came from.

Documents the assistant writes: Word, PDF and exact-layout sheets

Ask the assistant to write a document and it saves a real file in your files, with a preview and versions like any other document.

There are two kinds:

• Reports, minutes, memos and other flowing text. The assistant writes the text and saves it as a Word document (the default, editable), a PDF or a Markdown file. Say which format you want.
• Exact-layout sheets, for when the layout itself is the point: a client's form or checklist reproduced from its original, a one-page test sheet, a certificate or cover with a logo in a corner, tables with merged cells and shaded bands. The assistant lays the page out itself (it writes it as an HTML page) and saves it as a PDF. Nothing is added to such a page — no cover, no logo, no footer — unless you ask for it.

Getting a good exact-layout sheet:

• Attach the original form, or name the document to copy. The page is A4 unless you say otherwise.
• Logos and pictures must be images in your files or attached to your message — the assistant cannot take a picture from a web address. It can cut a logo or a figure out of a page of one of your documents, save it as an image, and place it on the sheet.
• The assistant looks at the finished pages and the page count itself. Ask for corrections in plain words ("move the logo to the top right", "keep it on one page"): each correction saves a new version of the same file, not another file.

Where it goes: the folder you name, or your own Chat files when you name none. Find your Chat files in Documentation, under the Chat files scope.

Creating a document is a change, so it follows your approval mode, and it needs the documents.create permission in a project. The Commissioning, Project Management and My Workspace assistants can write documents.

How to create a project from the chat

You can set up a new project by asking for it — also when you have no project yet and are chatting from My Workspace.

1. Open the chat and ask, for example: "Create a project called Riverside Data Center for Acme".
2. Give the two things the assistant needs: the project name and the client company. The project code is optional — a short one is made from the name if you do not give one. You can also ask which project templates exist and pick one.
3. An "Approval required" card appears with the project name, code and client. Creating a project always asks, whatever your approval mode. Click Approve.
4. A "Project created" card appears. Click Open to switch to the new project.

You become the owner of the new project. If the client company is not in the company directory yet, the assistant tells you and asks before adding it. This works with the Commissioning, Full, Project Management, My Workspace and General assistants.

Thinking mode: deeper answers for harder questions

The toolbar under the message box has a thinking-mode menu with four settings:

• Non-thinking — fast, direct responses.
• Thinking · Low — brief reasoning; quick answers with a sanity check. This is the recommended setting.
• Thinking · Medium — balanced reasoning for complex tasks.
• Thinking · High — deepest reasoning; thorough analysis for hard problems, with a slower reply.

Raise it for one hard question and lower it again afterwards. Under an answer, "Thought for … s" shows how long the assistant reasoned; click it to show or hide the reasoning.

Mention members with @ and assets with #

While typing in a project chat:

• Type @ to get a list of project members and mention one.
• Type # to get a list of assets and reference one directly.

Use the arrow keys and Enter, or click, to pick from the list; Escape closes it. A mention gives the assistant an exact person or asset to work with, so it does not have to guess.

Copy, rate and regenerate an answer

Under the assistant's last message of a turn you find:

• Copy — copies the answer text.
• Like / Dislike — rate the answer.
• Regenerate response — asks for a fresh answer to the same request. Only the last turn can be regenerated.

If the previous answer already made changes, a "Redo this request?" dialog tells you how many completed actions will be undone first; click "Undo & redo" to continue. When a turn includes an action that cannot be safely undone, Regenerate is not available for it.

Approvals: control what the assistant may change

You decide how much the assistant can do on its own. The approval mode button is in the toolbar under the message box, on the right, with a shield icon and the current mode's name. Click it to choose:

• Auto — the assistant runs all actions without asking. This is the default.
• Ask for critical — only critical actions (permanent deletions, member removal) ask first.
• Ask for edits — every action that changes project data asks first.

The mode is remembered per chat. Reading data never requires approval — only changes do.

A few actions ask in every mode, Auto included: creating a project, adding a company to the company directory, deleting a planner activity, and anything a connected service sends on your behalf, such as an email or a meeting invitation. For those, the card shows the full details — recipients, subject and message — so you can read exactly what will be sent. Connected services are set up in Settings → Connections.

Deleting a planner activity: the assistant can remove an activity from the planner when you ask, in the Commissioning Assistant. It always asks first — the card names the activity and warns that everything under it goes too — and it needs the planner.delete permission, like deleting it yourself.

How to answer an approval card

When the assistant needs your go-ahead, it pauses and an "Approval required" card appears in the conversation. It lists each action the assistant wants to run, with a checkbox per row.

1. Read each row. The title says what will happen; the line below shows the details.
2. Untick any action you do not want.
3. Click Approve to run the ticked actions (the button reads "Approve 2" when only some are ticked), or Decline all to run none.
4. Optional: before approving, tick "Don't ask again for … in this chat" so that kind of action stops asking for the rest of the conversation. Critical actions never offer this.

Declined actions show as "Declined" and the assistant carries on without them. If you send a new message instead of answering the card, the waiting actions are dropped and do not run.

If you are not looking at the chat when the assistant stops for approval, you get a notification: "The assistant is waiting for your approval".

Long tasks keep running in the background

The assistant's work is not tied to your browser tab. If you close the tab, open another chat, navigate away, or lose connection while it is working, it keeps going and the finished answer is waiting when you come back.

You can have the assistant working in up to 3 chats at the same time — start a long task in one chat, then open a new chat and ask something else. Each chat works on one request at a time: sending a new message in a chat that is still working stops the current answer and starts on the new message.

A small cluster of dots shows this work, one dot per chat:

• Pulsing dot — the assistant is working in that chat.
• Green dot — "Finished — click to view": an answer you have not seen yet.
• Amber dot — "Waiting for your approval".
• Red dot — "Stopped with a problem — click to view".

Point at a dot to see the chat's name, and click it to jump to that chat. The dots only show chats of the project you are in. Avoid asking two chats to change the same documents at once.

When an answer finishes in a chat you are not looking at, you also get a notification: "The assistant finished answering". If no Cx Assistant window is open anywhere, it reaches your phone as a push notification (when you use the mobile app with notifications on). You can switch these notifications off in Settings → Notifications.

Long conversations: the context meter and summaries

Every chat has a limit on how much the assistant can keep in its working memory at once. The small ring next to the message box shows how much of that memory the conversation is using — point at it (or tap it) to see the numbers. It fills as the conversation grows, and turns from green to amber to red as the limit gets closer.

When a conversation gets close to the limit, the assistant summarizes the older part of it and continues from that summary, so you can keep working in the same chat. This can take a minute or two, and the status line says "Summarizing earlier conversation…" while it happens. A thin line saying "Earlier conversation summarized" then appears in the transcript, and the ring drops. Nothing you wrote is removed — every message stays above the line for you to read; only what the assistant keeps in memory changes. Click "View summary" on the line to read what the assistant continues from.

You can also ask for a summary yourself, for example before starting a new part of a long job: point at the ring (or tap it) and press "Summarize conversation" in the panel that opens. The chat must be idle (not answering, and not waiting for an approval); while it is busy the button is greyed out and the panel says why. The ring also moves during long jobs: it updates after each step the assistant takes, not only when the final answer arrives. The whole conversation is summarized, and the assistant's next answer starts from that summary.

Summaries keep record ids, decisions, open tasks and the exact next step, but fine detail from early in the conversation may be condensed. For a new topic, a new chat is still the cleanest start.

Interactive cards in answers

Some answers include interactive cards instead of plain text:

• Add-issue card: a prepared issue you can review and confirm, creating it right from the chat.
• Dashboard card and metric chart card: live charts and metrics rendered inline, so the assistant can show you data, not just describe it.

These cards work like miniature versions of the real pages and respect the same permissions.

Manage your chat history

On the full chat page, the "Chats" sidebar lists all your conversations; in the drawer, click the history button at the top to slide the same list in.

• Click one to reopen it — the assistant remembers the whole conversation.
• Start a new chat with the "New chat" button.
• Search: type at least 2 characters in "Search chats…". Results show the matching line of the conversation, marked "You:" or "AI:".
• Rename or delete a chat with the icons on its entry in the list. Deleting asks for confirmation and is permanent.

The app remembers your last chat and returns to it. Chats are per project: switching project starts you fresh in that project's context, and your conversations stay separated per project.

Voice input: dictate your message

Click the microphone in the composer to dictate your message instead of typing — your speech is transcribed into the input box, where you can edit it before sending.

Voice input depends on your browser's speech support, so it may not be available in every browser. If you don't see the microphone, try a different browser.

Quick prompts on the welcome screen

When you open a new chat, the welcome screen offers quick prompts — ready-made questions you can send with one click. They are page-aware: the suggestions match what you are likely to ask from where you are.

Use them as starting points to discover what the assistant can do, then refine with your own follow-ups.

Teach the assistant with memories

You can give the assistant standing context it remembers across conversations — preferences, conventions, facts about how you work. Manage these in Settings under the Memories tab: view, add, edit, and delete memories in two scopes, Global (follow you everywhere) and Project (apply in one project).

Team-wide memories shared by the whole project are managed by project admins in the project's settings.

Chat and project scope: which project am I talking about

The assistant works in the context of your active project — its answers, searches, and actions target that project. Switching project restarts the chat session so contexts never mix.

You can also chat with no project at all (for example in personal scope): the assistant still works, and any files you upload land in the chat folder of your personal My Workspace instead of a project.

The assistant respects your permissions

There is no special permission needed to use the chat — every signed-in user can. But everything the assistant does on your behalf is checked against YOUR project permissions on the server:

• It can only read data you can read.
• It can only change what you could change yourself.

So the assistant never lets anyone see or do more than their role allows. If it reports that it can't access something, that usually reflects your own permissions — ask a project admin if you need more access.

Assistant problems — troubleshooting

• The assistant says it can't find or open something: it is limited to your own permissions — check that you can see the item yourself, and for documents check the Processing status in Documentation's preview pane (unprocessed files can't be read yet).
• A run seems stuck: check the run indicator dots — long tasks keep running in the background; you can also stop the response with the stop button in the composer.
• Actions keep asking for approval: your approval mode may be "Ask for edits" — switch the approval mode button under the message box to Auto or "Ask for critical", or tick "Don't ask again … in this chat" on the approval card. A few actions (creating a project, deleting a planner activity, sending an email through a connected service) always ask.
• The assistant seems to be waiting and nothing happens: look for an "Approval required" card in the conversation, or an amber dot — it is waiting for your answer.
• You cannot find the assistant picker: it only shows on a new, empty chat, and only when more than one assistant is available to you. Start a new chat to switch.
• A file is refused: check the format and the limits — 10 files per message, 60 MB per file, 200 MB for a video, one video per message.
• No microphone button: voice input depends on browser support — try another browser.
• Wrong project context: check the active project in the project switcher; chat sessions are per project.
• Uploaded a file with no project active: it went to your personal My Workspace chat folder, not a project.