AI Agent Integrations: Connect Omni to Every Tool Your Company Uses

Key Takeaways

  • Omni connects your company’s tools to one AI agent, giving it context from multiple sources.
  • Live integration documentation covers connectors such as HubSpot, Linear, ClickUp, Fireflies, Slack, GitHub, Microsoft 365, and more.
  • Permission-aware access helps ensure the AI agent only uses information users are authorized to access.
  • Cross-tool context lets Omni answer business questions using information from multiple connected applications.
  • Open-source code allows IT and security teams to inspect the connector implementation for review and compliance requirements.

Want to evaluate Omni directly? View Omni on GitHub or contact us to book a demo.

One AI agent, all your company’s tools

Omni is an open-source, self-hosted AI agent for company-wide work.

Its AI agent integrations connect the tools your company already uses. It has been built like this so that employees can:

  • search information,
  • gather context,
  • analyze data, and
  • carry out supported actions from one place.

Instead of creating a separate automation for every application, Omni brings information from connected tools into one shared context. This lets an agent work across systems when a task requires information from more than one source.

For example, an employee can:

  • Use meeting information from Fireflies and customer records from HubSpot to prepare for a follow-up.
  • Check an issue in Linear and find related information from connected company sources.
  • Review ClickUp tasks alongside information from documents, email, or other connected tools.

The result is one AI agent that can work with context from multiple company systems.

Browse Omni’s AI agent integrations

Omni’s integration directory focuses on connectors with live documentation. Each documented connector explains what information Omni can access. It also explains how to configure the connection.

Four integrations have full guides:

HubSpot

Omni can index HubSpot contacts, companies, deals, tickets, calls, emails, meetings, notes, and tasks. This gives the AI agent access to relevant CRM information for search and context.

Linear

The Linear connector indexes issues, projects, project updates, and documents. It can provide issue status, descriptions, comments, project information, and other documented workspace content.

ClickUp

Omni can index ClickUp tasks, comments, and documents. Teams can use this information to give the AI agent context about tasks, priorities, status, and project documentation.

Fireflies

The Fireflies connector indexes meeting transcripts, summaries, action items, and meeting metadata. This lets Omni use information from recorded meetings as part of its company context.

Other documented connectors

The current Omni documentation also includes these live connector pages:

  • Google Workspace: Drive files, Docs, Sheets, Slides, Gmail, and Google Chat content.
  • Google Ads: Account structure, campaigns, ad groups, keywords, assets, and related advertising data.
  • Microsoft 365: OneDrive, SharePoint, Outlook email, Calendar events, and Teams chats.
  • Atlassian: Confluence content and Jira issues.
  • Slack: Messages, threads, and file attachments.
  • GitHub: Repositories, issues, pull requests, and discussions.
  • Salesforce: Accounts, contacts, opportunities, leads, cases, tasks, and supported actions.
  • Darwinbox: Employee directory, organization data, holidays, ATS jobs, and HR workflow information.
  • Notion: Pages, databases, and workspace content.
  • IMAP Email: Email messages and attachments.
  • Nextcloud: Files and documents from a Nextcloud instance.
  • Paperless-ngx: OCR text, document metadata, tags, correspondents, custom fields, and notes.
  • Filesystem: Local files and directories.
  • Web: Public websites that Omni can crawl and index.
  • Windshift: Work items, descriptions, comments, and supported MCP-powered actions.

Omni also provides Python and TypeScript SDKs for building company-specific connectors when a required source is not in the documented connector catalog.

How enterprise AI integrations work in Omni

How Omni Connectors Work

Omni connectors act as a bridge between company tools and the AI agent. The connection follows three steps:

  1. First, the connector authenticates with a source.
  2. Second, it syncs accessible information.
  3. Then, it makes that information available for search and AI-powered answers.

The exact setup depends on the source. Connectors can use API keys, OAuth credentials, or service accounts. After the initial sync, supported connectors can update the indexed information as the source changes.

Permissions are an important part of this process. Omni uses the permissions granted by the workspace and the connected source. When a source supports permission inheritance, users only see information they already have access to in the original application. The AI assistant also applies the same permission filtering when retrieving information.

Because Omni is open source under the Apache 2.0 license, its core code is publicly inspectable. The public repository also contains the connector implementations. This can give IT and security teams an additional way to review connector behavior.

Learn more about the Omni open-source project.

Built on real connector docs, not a logo wall

Omni’s integration directory focuses on connectors that have supporting documentation. That makes it easier to evaluate what a connector actually reads, how authentication works, and what limitations apply.

This distinction matters because an application appearing in a product logo list does not necessarily mean that its integration is fully documented or available in the current product.

Omni’s documentation provides a specific connector page for each source included in the current directory. The integration list should therefore be evaluated against the live documentation rather than a static list of logos.

Connect your AI agent to company tools: examples

You can connect AI agent to company tools when a business question depends on information stored across different systems.

1. Turn meeting notes into customer context

A sales call is recorded in Fireflies. Omni can use the meeting transcript, summary, and action items alongside relevant HubSpot customer information.

This gives the agent context from both the meeting and CRM rather than relying on either source alone.

2. Investigate an issue across company information

A product or engineering team can ask about an issue recorded in Linear and use information from other connected company sources to understand the surrounding context.

For example, Omni can use the Linear issue details together with relevant documents or messages available through connected sources.

3. Review tasks with supporting information

A team can use ClickUp for tasks and connected sources for supporting documents or company knowledge. Omni can bring information from these systems together when answering a question about project work.

4. Answer one question using multiple sources

Some business questions do not belong to a single application. An employee may need information from a CRM, project management tool, meeting records, and company documents.

With multiple connectors enabled, Omni can search across the relevant connected sources and use the available context to answer the question.

FAQs

Do I connect each application once, or does every employee connect their own account?

Setup is connector-specific. Some sources use an administrator-configured API key or app token to sync data; others use user-level authorization. Action tools may also require a separate per-user OAuth connection even when a search source is already configured. Check each connector’s setup guide rather than assuming that one shared credential covers indexing, employee access, and actions.

Do all connectors inherit source permissions?

No. Omni uses the permission metadata a connector supplies, and that behavior varies by source and content type. For example, the current HubSpot and Fireflies connectors make indexed content searchable by all Omni users. Linear team filters restrict issue fetching, not workspace-wide projects and documents, and ClickUp Docs use workspace-level permissions. Review each connector’s visibility rules before syncing sensitive information.

What happens when I connect a tool that can read data but cannot take actions?

Its indexed content can still support search and cross-tool agent answers. A read-only connection does not let Omni edit the source application or turn every answer into an automated workflow. Where action tools are supported, they require their own configuration and authorization. ClickUp, for example, separates API-token indexing from per-user MCP OAuth actions.

How fresh is the information in Omni’s search index?

Indexed answers depend on the last successful sync, not a guarantee of live access to every application. Supported connectors perform an initial sync and subsequent updates, with behavior varying by source. Check the source’s sync status, last sync time, and errors in Settings → Integrations when recent content is missing. A separately authorized live tool is different from a synced search index.

Do integrations work on self-hosted Omni?

Yes. Run the required connector services as part of your Omni deployment, then configure each source’s credentials and start its initial sync. The free Community version includes integrations; paid self-hosted licenses add enterprise capabilities. Self-hosting does not bypass a source application’s API requirements, rate limits, or access rules.

What if my company uses a tool that is not listed here?

Check the connector documentation first: suites such as Google Workspace and Microsoft 365 cover multiple applications under one connector. For an unsupported system, Omni provides connector SDKs for building a custom integration. The Enterprise offering also includes custom connectors; contact the Omni team to discuss requirements rather than assuming any unlisted application is supported automatically.