Shared knowledge base
Upload files, write documents, and organize the knowledge your workspace uses to answer questions and investigate problems.
Start with a question. Connect another tool. Hand over a recurring task. Explore the capabilities your team can put to work.
26 features
Features depend on your workspace setup, connected sources, and permissions.
No matching features. Try a different search or clear the filters.
Give your team’s knowledge a place to grow.
Upload files, write documents, and organize the knowledge your workspace uses to answer questions and investigate problems.
Review how a document has changed and restore an earlier version when you need it.
Explore connections between documents and generated system knowledge. Follow relationships to see what belongs together.
Keep the connections in your knowledge graph useful. Dream Cycles strengthen well-used links, discover related knowledge, and prune weak links on a schedule.
Administrators can set the schedule, pause cycles, and revert recorded automatic changes. Uncertain connections become questions; proposed directional relationships, such as dependencies, require confirmation. Pinned links are protected from decay and pruning.
Answer questions about missing knowledge or uncertain connections. Confirm what is right, pin a relationship worth remembering, or reject a suggestion.
Dream Cycles maintain knowledge relationships. They do not train the underlying AI model or automatically fix your systems.
Ask the question, then follow it where it leads.
Ask in plain language. Vedin searches the knowledge available in your workspace and can read the documents behind an answer.
Explore connected business data and turn results into summaries, tables, and charts.
Results depend on available data. Review gaps and assumptions; a generated analysis is not a validated forecast.
Ask in Vedin or Google Chat, then continue the conversation using the earlier thread as context.
A connected chat channel is not automatic ingestion of all chat history.
Understand the systems you have and the changes ahead.
Ask about indexed repositories, find definitions and references, and trace call chains without piecing everything together by hand.
Use generated repository documentation and cross-repository relationships to understand how services fit together.
Coverage depends on connected repositories and successful indexing.
Review pull requests with repository context and your documented conventions in view. Use proposed fixes as a starting point for your review.
Review and test proposed changes before merging or deploying them.
Give compatible coding assistants access to your Vedin knowledge. Search and read documents from tools such as Claude Code, Codex, and Cursor.
Vedin’s knowledge MCP server is read-only. Access requires an enabled endpoint and a personal token; it stays within the token owner’s permitted knowledge scope. Tokens can expire or be revoked.
Hand over the repeat work. Come back to the result.
Describe recurring work and choose a schedule. Have Vedin prepare reports, perform a triage sweep, or carry out a task using connected tools.
Each run has a recorded status and result. Available tools, permissions, and usage limits still apply.
Set up repeatable pass/fail checks for supported conditions, such as an HTTP response, certificate expiry, or database row count.
Checks use a saved execution plan. Depending on configuration, failures can trigger adaptation or an investigation.
See what ran, what happened, and the recorded usage. Pause recurring work when it is no longer needed.
Start from a problem or a supported alert webhook. Vedin gathers evidence with connected tools, records findings, and works toward a next step.
Connect a supported alert source to start investigations automatically. Review recorded evidence before applying a proposed fix.
Use structured runbook documents to guide repeatable work through Vedin’s tools.
Execution depends on supported runbook steps, enabled tools, and applicable action policies.
Bring the systems your team already relies on.
Connect supported document sources, chat, databases, code hosts, and cloud systems. Use their knowledge and tools in the same workflow.
The integration catalogue distinguishes available connections from those coming soon.
Bring tools into Vedin from an internal or third-party Model Context Protocol server. Configure the server, discover its tools, and test the connection.
Vedin calls tools provided by the connected server. Available actions depend on the server’s tools, authentication, and configured permissions.
Connect internal APIs, scripts, and other supported tool types when a built-in integration does not cover your workflow.
Configure what the tool does, test it, and assign an approver when a person must review its use.
Reference named values in tool configurations and task prompts. Keep environment-specific settings separate from the work you want to run.
Use an infrastructure agent to connect supported systems inside your network, including private databases and infrastructure tools.
The agent provides access to supported systems. It does not self-host the complete Vedin service or its AI models.
Keep people involved where their judgment matters.
Review actions that require your approval before they run. See the proposed action and its context, then approve or reject it, with the decision recorded.
Approval requirements depend on the tool’s action policy, approver, and connection permissions.
When Vedin needs information from you, answer in the workflow so it can continue with the missing context.
Review recorded findings, tool calls, outputs, and approval decisions. See what Vedin tried and where it needs more help.
Manage who can invite people, maintain knowledge, configure connections, and manage tools in your organization’s workspace.
Access to a connection does not bypass its credentials or execution policy.
Walk through the product, or start with the work your team needs help with.