Takibi Base · A collaborative knowledge base for AI agents

Takibi Base is a shared knowledge base designed for AI agents, enabling teams to organize documents and control access for each agent. It provides precise passages with citations when agents query, streamlining the process of information retrieval and management. Users can upload documents, manage permissions, and ensure that agents have access to the necessary information while maintaining document integrity.

1. Document Management

   Users can upload files into projects and folders, with Takibi converting them to Markdown for easy searching. Original files can be downloaded or text extracted again.

2. Profile Access Control

   Each Profile holds access permissions for its keys, allowing users to specify which folders an agent can read. This ensures that sensitive information is only accessible to authorized agents.

3. Query and Citation System

   When an agent queries the system, Takibi returns exact passages with citations, providing evidence support scores to validate the information retrieved.

4. Task Assignment

   Users can create tasks within projects, assign them to agents, and review their work. This feature enhances collaboration and ensures accountability.

5. Note-taking Functionality

   Agents can leave notes on their findings, which can be reviewed and utilized for future tasks. Notes have a 30-day lifetime and are private to each Profile.

6. Export Options

   Users can export their knowledge base and original files in Open Knowledge Format (OKF), allowing for external storage and backup.

1. Create your workspace by signing up and logging into Takibi Base.
2. Upload your documents into designated projects and folders for organization.
3. Set up Profiles to control access permissions for different AI agents.
4. Use the CLI to query information and receive exact passages with citations.
5. Regularly review and manage notes left by agents to maintain an organized knowledge base.

What is Takibi Base?

Takibi Base is a shared knowledge base for AI agents that allows users to organize documents and control what each agent can access.

How does Takibi Base work?

Users upload documents, which are then indexed and made searchable by AI agents. Agents can query the system to retrieve specific passages along with citations.

Can I control which documents each AI agent can access?

Yes, access is managed through Profiles that specify which projects or folders an agent can read.

Does Takibi Base generate answers?

No, Takibi Base does not generate answers; it only returns exact passages from documents.

What types of documents can be uploaded?

Takibi supports PDF, Word, Markdown, and plain text documents, provided they have extractable text.

Takibi offers a 7-day trial, after which the pricing is as follows:

• Ember Plan: 18 EUR per month for 2 hired Profiles.

• Campfire Plan: 52 EUR per month for 8 hired Profiles.

Prices are exclusive of VAT.

1. Takibi Base effectively organizes documents and allows for precise access control, making it suitable for teams that rely on AI agents for information retrieval.
2. The citation system enhances the credibility of the information provided, which is crucial for decision-making processes.
3. However, the reliance on user-uploaded documents means that the quality of information is contingent on user diligence in maintaining and updating the knowledge base.
4. The note-taking feature is beneficial but may require users to actively manage and review notes to ensure they remain relevant.
5. Overall, Takibi Base is a robust tool for managing knowledge for AI agents, but users must be proactive in document management and access control to maximize its effectiveness.