Chatbot integrations that connect your content and your channels
matram.ai splits its chatbot integrations into two jobs: connectors that pull your knowledge in, and channels that push your bot out. Connect a source, sync it, and the same trained bot answers on your website, in Slack, and inside your help desk.
Two kinds of chatbot integrations, one trained bot
matram.ai has two integration families: data source connectors that feed training content, and channel integrations that deploy the bot. Both point at the same knowledge base.
Every integration in matram.ai does one of two things. A data source connector reads your content (a Notion workspace, a Google Drive folder, a Confluence space) and turns it into training material. A channel integration takes the bot you trained on that material and places it where people already ask questions. You set up the first group once, and the second group inherits everything.
This separation matters because your content and your reach change independently. You might add a Freshdesk help center as a source this month and add Slack as a channel next month, without retraining or rebuilding anything. The retrieval engine reads from one indexed knowledge base, so a fact you import through Google Drive shows up whether a visitor asks on your website widget or a teammate asks in Google Chat.
If you are still deciding what the bot should be able to do before wiring up sources, the features overview walks through answering modes, human hand-off, and analytics. This page focuses on the plumbing: what connects, how it connects, and how it stays secure.
Data source connectors that feed training content
Connect a knowledge source and matram.ai cleans, chunks, embeds, and indexes it so answers cite your own material.
Notion
Import pages and databases from a Notion workspace. Good for internal wikis, product docs, and policy pages that already live in Notion.
Google Drive
Pull documents and folders straight from Drive. Useful when your team drafts help articles and specs as Google Docs.
Dropbox
Connect a Dropbox folder of PDFs, docs, and exports so shared files become answerable content.
OneDrive
Bring in files from OneDrive for teams standardized on Microsoft 365 for day-to-day documents.
SharePoint
Index SharePoint document libraries and pages, a common home for internal knowledge in larger organizations.
Box
Connect Box folders so governed enterprise content can train the bot without leaving your content workflow.
Confluence
Import Confluence spaces and pages. This is often the single biggest source of support and engineering knowledge.
Gitbook
Pull structured product and developer documentation from Gitbook into the knowledge base.
Zendesk Help Center
Sync your existing Zendesk help center articles so the bot answers from the same content your agents rely on.
Freshdesk
Import Freshdesk knowledge base articles to reuse support content you have already written.
YouTube
Add YouTube videos so transcripts and explanations in your videos become searchable, citable answers.
Website, sitemap, files, and text
Beyond connectors, crawl a website URL, import an XML sitemap, upload PDF or DOCX files up to 10MB, or paste raw text directly.
Channel integrations that deploy the bot
Once trained, the same bot goes live across eight channels, with WhatsApp and HubSpot marked coming soon.
Website widget
Add the chat widget with one line of JavaScript. Customize color, position, avatar, welcome message, quick prompts, and more, then remove branding on Standard and higher plans.
Slack
Bring the bot into Slack so your team can query your knowledge base without leaving their workspace.
Messenger
Answer Facebook Messenger conversations with the same grounded, cited responses your website visitors get.
Crisp
Deploy inside Crisp so incoming chats get an instant first answer before an agent steps in.
Freshchat
Connect to Freshchat and let the bot handle common questions inside your existing messaging setup.
Zendesk
Run the bot in Zendesk alongside your agents. See how it compares on the Zendesk alternative page.
Zoho SalesIQ
Add the bot to Zoho SalesIQ to answer and qualify visitors on sites already using SalesIQ.
Google Chat
Put the bot in Google Chat so teams on Google Workspace can ask questions in a space they already use.
WhatsApp (coming soon)
WhatsApp deployment is on the roadmap and marked coming soon.
HubSpot (coming soon)
A HubSpot integration is planned and marked coming soon.
How connecting an integration works
Connecting is per workspace and takes a few clicks: authorize the source, pick what to sync, and let matram.ai index it.
- 1
Choose the connector
From your workspace, pick the source you want to add (for example Confluence or Google Drive). Each connection belongs to that one workspace, so a source you add in one workspace never leaks into another.
- 2
Authorize the connection
Connect the source and matram.ai stores an access token so it can read your content. The token is encrypted at rest and scoped to that one workspace, and it exists only so the connector can sync material into the knowledge base.
- 3
Select the content to sync
Point the connector at the specific spaces, folders, or articles you want. You control the scope instead of importing everything by default.
- 4
Let matram.ai clean, chunk, and index
Imported content is cleaned, split into chunks, embedded with pgvector, and indexed. Retrieval-augmented generation then grounds every answer in that content and adds source citations.
- 5
Set the auto-sync cadence
Auto-sync re-crawls your sources on a cadence that scales with your plan, daily, weekly, or monthly, so answers stay current as your documents change. No manual re-upload.
Security of connectors
Connector tokens are encrypted at rest, every connection is isolated to its workspace, and your data never trains AI models.
What teams do with these integrations
The point of chatbot integrations is fewer repeat questions and more captured demand, wherever people ask.
Deflect repeat support tickets
Sync your Zendesk or Freshdesk help center as a source, deploy to your website widget and Crisp, and let the bot answer common questions with citations before a ticket is ever filed. Escalation to a live agent stays one click away.
Answer from internal docs in Slack
Connect Confluence, Notion, or Google Drive, then add the bot to Slack or Google Chat. Your team asks in plain language and gets grounded answers instead of digging through wikis.
Capture leads across channels
Deploy to Messenger, Zoho SalesIQ, and your website, then turn on the lead capture form and in-chat qualification. Every channel feeds the same leads list. See the details on the lead generation page.
Keep video knowledge answerable
Add YouTube tutorials as a source so the explanations in your videos become searchable text answers, not content locked inside a player.
One knowledge base, one deployment, every plan
All chatbot integrations work against a single indexed knowledge base, and channel access scales with your plan.
Because sources and channels share one knowledge base, you never maintain parallel copies of your content per channel. Train once, deploy everywhere the bot is available. Answers stay consistent whether someone reaches you through the website widget or Google Chat, and they carry source citations so people can verify where a fact came from.
Plans set how much you can connect and run. Basic at $29/mo includes 1 chatbot, 2,000 messages, and 3 sources. Standard at $69/mo raises that to 5 chatbots, 5,000 messages, all channels, the public API, and branding removal. Pro at $199/mo covers 20 chatbots, 20,000 messages, and priority support. Enterprise is quoted rather than published, for custom volume with invoiced billing, so talk to sales if you need more than Pro. Every published plan starts with a 7-day free trial and no credit card.
Not sure what the return looks like before you connect anything? Estimate it with the free chatbot ROI calculator, then start a trial and wire up your first source.
How a connection becomes an answer
Connect a source once. Matram keeps it in sync and answers from it, with the page cited on every reply.
Integration setup guides
Step-by-step guides for connecting matram.ai to the tools you already run.
Every integration
Data sources for training
Bring content into a bot from the tools where it already lives.
Google Drive
Seamlessly access and sync chatbot training data stored in Google Drive.
Dropbox
Integrate with Dropbox to use stored documents and files for training.
OneDrive
Cloud-based storage of chatbot training materials with easy access.
SharePoint
Organize and share chatbot training documents within a team.
Zendesk
Access help-center content as a rich source of training information.
Gitbook
Connect to structured documentation and knowledge bases.
Box
Secure cloud storage of training data, safe and organized.
Notion
Train on notes and organized data from your Notion workspace.
Freshdesk
SoonAccess help-center articles to enrich the chatbot's knowledge base.
Confluence
SoonUse comprehensive documentation and collaborative content.
YouTube
Train using YouTube videos, playlists, and channels.
Channels for deploying
Put the same assistant to work in the channels your customers use.
Google Chat
Let customers get instant support in their preferred messaging app.
Messenger
Handle customer queries in real-time on Facebook Messenger.
Crisp
Automatically answer common queries and free up your team.
Slack
Power a shared customer-service channel with your users.
Freshchat
Resolve common issues for a smooth customer-service experience.
Zendesk
Improve ticket response times and let agents focus on high-touch issues.
Zoho SalesIQ
Engage visitors, answer queries, and convert leads via live chat.
Offer immediate, intelligent responses in WhatsApp.
HubSpot
SoonChat with users and escalate to a human when needed.
Frequently asked questions
Connect a source, deploy a channel, in one afternoon
Book a demo with no credit card. Wire up Notion or your help center, drop the widget on your site, and add Slack when you are ready.