Available integrations
| Integration | Type | Description |
|---|---|---|
| Messaging | Connect your agent to WhatsApp Business | |
| Slack | Messaging | Respond to direct messages and @mentions in Slack |
| WordPress | Website | Embed your agent in WordPress with a plugin |
| Zapier | Automation | Send leads and events to 7,000+ apps |
| Pabbly Connect | Automation | Alternative to Zapier for workflow automation |
| Webhooks | Developer | Send events to your own HTTP endpoints |
| Google Drive | Knowledge | Import documents, spreadsheets, and slides |
| Notion | Knowledge | Import pages and database entries |
| OpenAI | AI Provider | Use your own OpenAI API key (BYOK) |
| OpenRouter | AI Provider | Access models from Anthropic, Google, Meta, xAI, and more |
| Groq | AI Provider | Ultra-fast inference for supported models |
Messaging integrations
Connect your agent to messaging platforms so customers can interact outside of your website:- WhatsApp — Connect via Meta’s Embedded Signup flow. Customers message your WhatsApp number and your agent responds automatically.
- Slack — Add your agent to a Slack workspace. Team members can DM or @mention the agent in channels.
Knowledge integrations
Google Drive and Notion connect as content sources for your Knowledge Base. Connect them in Integrations → Hub, then import documents and pages from the Knowledge Base’s Add Resource tab. See the Google Drive and Notion setup steps.Automation integrations
Send agent events (leads, conversations, messages) to external tools:- Zapier — Build automated workflows with 7,000+ apps using a dedicated Zapier API key
- Pabbly Connect — Similar workflow automation using a dedicated Pabbly API key
- Webhooks — Send events to any HTTPS endpoint with signature verification
AI provider integrations
On BYOK (Bring Your Own Key) plans, you can connect your own API keys in Integrations → Hub:- OpenAI — Required for embedding models. Also provides access to GPT models.
- OpenRouter — Access models from Anthropic, Google, Meta, xAI, MistralAI, Cohere, Qwen, and DeepSeek
- Groq — Ultra-fast inference for models like Llama and GPT-OSS