---
title: "The context layer for your company’s AI — In Parallel"
description: "In Parallel maintains shared context from meetings and relevant chats, so your AI tools can work with what’s happening now. Every piece of context knows its audience."
url: https://www.in-parallel.com/
---

# The context layer for your company’s AI.

In Parallel maintains shared context from meetings and relevant chats, so your AI tools can work with what’s happening now. Every piece of context knows its audience.

[Get started →](https://www.in-parallel.ai/)

[Book a demo →](https://www.in-parallel.com/demo)

Works in Slack, Teams, Claude, ChatGPT and Copilot

From €29 per user per month

Billed monthly, no commitment

Already use Claude or ChatGPT? [Let it set you up →](https://www.in-parallel.com/set-up-with-ai)

## Ask about your work from the tools you already use

Ask what’s changed, what needs attention or what to do next. Use your existing tools or ask directly in In Parallel. The same shared context is behind every conversation.

How it works?

## In Parallel listens where work happens — and captures every decision as it's made.

1. joins your meetings
2. reads your Slack & Teams threads
3. …and any other tool you need.

In Parallel gives your AI the context and deep understanding you need to do your work better.

Sales

Account memory

Updated from Salesforce · just now

Customer success

Customer history

Updated from Zoom · just now

Management team

Decisions & commitments

Updated from Google Meet · just now

## The right context gives the right answer

An AI with no context says it doesn’t know. An AI reading last month’s plan gives a confident, wrong answer, and someone acts on it. The context behind a question changes the answer. In Parallel keeps that context current, shows where each answer came from, and gives people and their AI tools the level of detail they’re allowed to access.

[How it stays current →](https://www.in-parallel.com/how-it-stays-current)

### Your company knows everything

Your company’s knowledge stays together. Choose what this project and its AI can access.

Company context

- Project status
- Decisions
- Risks
- Owners
- Update history
- Contract value
- Day rates
- Board memo (NDA)
- HR thread
- Other clients

0 of 10 sources shared

**AI needs your input**Claude can provide a template, but you still have to supply the facts.

Draft the latest Project Alder update for our customer, Elena.

No project sources attached

Claude

I don’t have the latest information about Project Alder. Here’s a template you can fill in: Status: [on track / at risk] Progress: [completed this week] Next steps: [owner and date] Blockers: [issue and impact] Share the recent project updates and I can turn this into a customer-ready draft.

Illustrative responses · Fictional project

## Share a decision without the details behind it

With a file, access is yes or no. Sensitive knowledge rarely moves that way. A board shares the outcome with the company without sharing who argued for it. In Parallel lets each share carry one of three levels of detail, from the same underlying record.

### Full

Members of the workspace where the meeting belongs

Decision

Delay the Nordic launch to Q2. Raised by Anna Virtanen, CFO, in the leadership weekly on 12 September, citing a 14% overrun in supplier contracts.

### Sources unnamed

A neighbouring team that needs the reasoning

Decision

Delay the Nordic launch to Q2. Raised in a leadership meeting, citing an overrun in supplier contracts.

### Conclusions only

The wider company, a partner, or an external agent

Decision

The Nordic launch moves to Q2.

### How In Parallel compares with other ways to give AI context

Where the approaches do the same job, and where they part.

Scroll to compare — In Parallel stays in view →

|  | A context layer built from working conversations. It keeps shared knowledge current and gives each audience the level of detail they’re allowed to access. | An intelligence layer over Microsoft 365. It reads the access lists, sensitivity labels and data loss prevention rules set in SharePoint, Purview and Entra, and applies them to every request an agent makes. | A shared workspace inside Claude. It uses documents and instructions added to a project as context for its chats. On Team and Enterprise plans, members can be given view or edit access. | A manually maintained collection of company knowledge in documents or Markdown files. People curate, update and share the content. Access controls depend on where the files live and how an assistant reads them. |
| --- | --- | --- | --- | --- |
| What it protects | Decisions, risks, actions, notes and transcripts from meetings, in any tool | Files, mail, chats and calendars in Microsoft 365 | Documents, files and instructions added to a Claude project | Whatever people write down and keep up to date in documents or Markdown files |
| Who sets the rules | Workspace owners and company admins, in the product | IT, in SharePoint, Purview and Entra | Project creators manage access; organization owners control sharing on Team and Enterprise | File or repository owners, using the host’s permissions |
| AI answers limited to what the person may see | Yes, filtered when the question is asked, in search, chat and connected assistants | Yes, inherited from Microsoft Graph | Project members can use its shared knowledge; access is set at project level | Depends on the host and retrieval setup; copied files need their own access controls |
| Acts as the user, never as an app | Yes, with tokens that can be narrowed to chosen workspaces | Yes | Project permissions apply inside Claude; external tool access depends on the connector | Depends on the integration: a user account or a shared service account |
| Partial disclosure | Three levels, plus masking of people | No | View or edit access to project contents; no documented summary-only sharing level | People create and maintain separate summaries or redacted copies |
| Attendance-based access | Joined-only notes | No | No documented attendance-based project access | Must be set up manually in the hosting system |
| Sensitivity labels | Planned: classified topics, rated as content is captured | Purview labels, applied by people or rules, inherited by derived insights | No documented project-level sensitivity labels | Depends on the host; labels written in Markdown do not enforce access |
| Read history | Shared reads visible to the owner; member reads planned | Full audit for compliance admins | Enterprise audit logs for organization owners | Depends on the host and integration; version history records edits, not necessarily reads |
| Retention | Transcript and recording retention per company; data subject requests supported | Purview retention policies | Custom conversation and project retention on Enterprise | Depends on the host; downloaded and copied files must be managed separately |
| Separation between customers | A separate database schema per company | Microsoft 365 tenant | Organization and project access controls; database isolation model not specified in project documentation | Depends on the storage and retrieval architecture |

Sources: [Microsoft Work IQ](https://learn.microsoft.com/en-us/microsoft-365/copilot/extensibility/work-iq/) · [Claude Projects](https://support.claude.com/en/articles/9519177-how-can-i-create-and-manage-projects) · [Claude audit logs](https://support.claude.com/en/articles/9970975-access-audit-logs) · [Claude retention](https://support.claude.com/en/articles/10440198-configure-custom-data-retention-controls-for-enterprise-plans).

Rent the model. Own the context.

## Switch AI tools without losing what your company knows

Your AI tools will change. Switching a model takes an afternoon; rebuilding two years of company context takes two years. Keep the context in a layer you own, and let every tool read from it.

[Four tests of ownership →](https://www.in-parallel.com/own-your-context)

- One context for Claude, ChatGPT, Copilot, Cursor and your own agents, over MCP
- AI agents need no seat, so changing tools never changes the bill
- Never used to train models, and kept in the EU

Your data, your rules

## Your company decides who and what can see its data

Your company controls its data and who can use it.

- EU hosted
- ISO 27001, ISO 42001 and SOC 2 Type II certified
- GDPR compliant
- In Parallel never trains on your data
- Answers cite the meeting, decision or page they came from
- Revoke a share and the next request returns nothing — for people and for AI
- Share the conclusion without the source

[Read about security →](https://www.in-parallel.com/security)

ISO 27001

ISO 42001

SOC 2 Type II

GDPR

Pricing

## Price the knowledge, not the people.

Every user adds Knowledge Units to one pool the whole company shares. Reads from your own AI tools and agents over MCP are free within fair use.

For teams under 50 users, getting started with workspaces

### Professional

€29per user / month

€23.20 / month, prepaid annually

**Knowledge Units**

120 per user / day

- Meeting notes, chat, pages and workspaces
- Your AI connected over MCP

For teams that share work with AI tools and agents

### Business

€69per user / month

€55.20 / month, prepaid annually

**Knowledge Units**

250 per user / day

- Everything in Professional
- Sharing: give people access to a workspace, with set access rights for what they can see
- Up to 5 app integrations
- Automations, when released

From 50 users, everyone in the company on one contract

### Enterprise

Custom

One annual price, invoiced in advance

- Everything in Business
- Unlimited app integrations
- SSO
- A named contact

[Talk to us →](https://www.in-parallel.com/demo)

One plan for the whole company: moving up moves everyone. Every user who signs in has a seat, including clients added to a workspace; AI agents need none.

[Get started →](https://www.in-parallel.ai/)

[Book a demo →](https://www.in-parallel.com/demo)

Billed monthly at list price, cancel any month. Prices exclude VAT.
Prepay a year and save 20%.

## A shared understanding, with clear boundaries.

Keep people and their AI tools informed as work changes. Control what each audience can see.

[Get started →](https://www.in-parallel.ai/)

[Book a demo →](https://www.in-parallel.com/demo)

20 minutes with a founder

From €29 per user per month · Billed monthly, no commitment
