Integrations
Kluco plugs into what the company already uses
Sources feed the Brain, AI models sit side by side in LLM Management, agents answer in the team's channels and any other AI reads the memory over MCP. This page lists only what exists in the product today.
Brain
Knowledge sources
Every source has a sync schedule and history. What comes in becomes an article with an owner, a review deadline and an audience set by the company.
Google Drive
Folders and documents from the company Drive.
OneDrive and SharePoint
Files from the company's Microsoft 365.
Dropbox
Dropbox folders and files.
Amazon S3
Documents stored in AWS buckets.
Google Cloud Storage
Documents stored in Google Cloud buckets.
Document upload
PDF, TXT, Markdown, Word, Excel and CSV, with folders and labels.
Support conversations
Resolved conversations, by channel and inbox, with personal data masked and a review before they become knowledge.
Webhook and API
Any system pushes content to the Brain, or the Brain pulls from a company API.
Brain
Recorded meetings
The minutes, or the minutes with the transcript, reach the Brain without the attendee list.
- Fathom
- Gemini in Google Meet
- Fireflies
- Zoom
- Read AI
- Granola
LLM Management
AI models
Providers sit side by side in the same console, with access profiles, limits and budgets. People pick the model in the Chat and can switch mid-conversation.
OpenAI
GPT models.
Anthropic
Claude models.
Google Gemini
Gemini models.
DeepSeek
DeepSeek models.
Grok
xAI models.
OpenRouter
Open models from many vendors through one account.
Agents
Agent channels
The agent answers where the team already talks, with the same rules, approvals and budget as in the console.
- Slack
- Microsoft Teams
- Google Chat
- Telegram
MCP and API
The open path
Memory and processes are not locked into Kluco: other AIs and other systems come in through open standards.
Brain MCP server
ChatGPT, Claude, Gemini and other tools that support MCP read the company memory, isolated per company.
Tools over MCP
Agents use tools from MCP servers, with a rule per tool: allow, ask for approval or block.
Use-case connector
People log what they do in the ChatGPT, Claude or Gemini of their own plan, and the company sees the aggregate by task and team.
Processes by webhook
Another system opens cases in Intelligent Automation with a key, without opening the same case twice.
External agents
AI routines that run outside Kluco report usage, cost and a summary, with a budget that warns or pauses.
Skills as SKILL.md
Imports and exports skills in the open Agent Skills format.
Wiki export
The whole base leaves in an open format, readable without Kluco.
FAQ
Integrations
Can we use the Kluco memory inside ChatGPT or Claude?
Yes. The Brain has its own MCP server, isolated per company. ChatGPT, Claude, Gemini and other tools that support MCP read the company memory through it, with the permissions set in the console.
Do we have to stop using ChatGPT or Claude to use Kluco?
No. Kluco sits outside any vendor: the team can use the Kluco Chat with several models, keep using ChatGPT, Claude or Gemini and read the memory over MCP, or both.
Does Kluco read recorded meetings?
Yes, from Fathom, Gemini in Google Meet, Fireflies, Zoom, Read AI and Granola. The company chooses whether only the minutes or the minutes with the transcript come in, and the attendee list is not stored.
The tool we use is not on the list. What now?
Any system can push content to the Brain by webhook, or the Brain pulls from a company API. For processes, another system opens cases by webhook with a key. If an integration decides the purchase, talk to the team in the demo.
Next step
See the integrations with your company's tools
In the demo we connect a real source and show the knowledge reaching the Brain and answering in the Chat and over MCP.