Skip to main content

What you can do with In Parallel's AI

Choose the right AI surface for a briefing, a page-aware question, a focused handoff, or work in a connected AI tool.

Written by Topi Järvinen

In Parallel gives you several ways to work with AI: Chat inside the product, focused handoffs with Run with AI, connected assistants over MCP, and background AI that maintains useful Workspace context.

Choose the right surface

Surface

Use it when

What makes it different

In Parallel Chat

You want a briefing, answer, or follow-up without leaving In Parallel

Ask can use the page on screen as context

Run with AI

You want to continue from one Workspace page or meeting in another AI tool

The handoff starts from the visible record

Connected assistant over MCP

You want Claude, ChatGPT, Cursor, or another client to work across In Parallel and other tools

The assistant retrieves authorized In Parallel data on demand

Background AI

You want meeting Notes, briefings, Topics, or maintained Workspace pages prepared from Workspace signals

It prepares context for people to review rather than waiting for a prompt

What to ask In Parallel Chat

Open Home for a full-page conversation, or select Ask on a page where that entry point is available to open Chat alongside what you are viewing.

Useful requests include:

  • “Give me the three things that need my attention today.”

  • “Summarize this meeting and separate decisions from unresolved questions.”

  • “What does this Workspace page say about the launch decision?”

  • “Prepare me for my next meeting with this Workspace.”

  • “Draft a follow-up email from these actions.”

  • “What evidence supports this risk?”

The Chat context chips show what the next message is grounded in. You can also add Workspaces, Topics, or documents explicitly.

A full-page In Parallel Chat response to a Workspace drift-report request, with a structured signal table and follow-up composer

See Use In Parallel Chat for the full workflow.

What to ask a connected assistant

Connect an AI assistant over MCP when the work should continue outside In Parallel or span more than one system. You do not need to know MCP tool names; describe the outcome and let the client choose the available tools.

You want to…

Ask something like…

Catch up after time away

“What changed in my Workspaces while I was out last week?”

Run a review

“Review the Launch Workspace: current risks, open decisions, and overdue actions.”

Triage work

“What is overdue across my Workspaces, and what should I address first?”

Turn discussion into structure

“Read last Thursday's planning meeting and propose goals from it.”

Work across tools

“Turn the approved launch decisions in In Parallel into Jira epics.”

Connect In Parallel and a delivery tool such as Jira to the same assistant when you want it to reconcile the two. Keep execution context in In Parallel and delivery detail in the delivery tool.

What remains under your control

  • Chat and connected assistants only see data you are authorized to access.

  • Context is rechecked when a turn runs; an old conversation does not preserve access you no longer have.

  • Generated summaries and recommendations can be wrong. Open the linked source before making an important decision.

  • External communications require you to review the recipients and message before sending.

  • A connected client's own privacy, approval, and retention settings still apply.

Related

Did this answer your question?