Clear boundaries.
Visible work.

Give Vedin useful work while keeping access, actions, and data handling in view. Review the controls and confirm the setup your organization needs.

Start a security review

Access you define.
Actions you can review.

Choose connections and tool permissions to match the work you want Vedin to do.

Organization and team access.

Your organization and team membership determine your access. Vedin searches knowledge available to your team, including information explicitly shared across teams.

Tools you choose to enable.

Administrators choose which tools Vedin can use. Disabled tools cannot run, and each connection’s permissions still apply.

A record you can inspect.

Review recorded tool calls, outputs, and approval decisions during an investigation. Use that record to understand the findings and decide what happens next.

Review the action
before it runs.

When a tool call requires approval, the investigation pauses for an authorized person’s decision. The work resumes after that decision, with the outcome recorded.

Approval requirements depend on each tool’s policy, its approver, and the connection’s permissions.

See controls in the product
A pending approval for a production-service restart, with a proposed command, success check, and rollback guidance.
A production restart waiting for approval. View full size ↗

Know how your
information is handled.

Storage, processing, and retention are different questions. Review each one for your configuration.

Customer-content encryption

Vedin supports organization-specific encryption for customer content, using AES-256-GCM for records and files and Google KMS to protect the organization keys.

Existing organizations may retain plaintext content until encryption rollout and plaintext retirement are complete. Confirm your organization’s encryption status during setup.

Connection credentials

Connection configuration and credentials are stored encrypted and resolved when a tool runs. Known secret fields and registered masked values are filtered from model context and recorded outputs.

Tool output can still contain sensitive data. Give connections only the permissions they need, and review the information their tools can return.

AI processing

Prompts, relevant knowledge, tool results, and conversation context may be sent to the configured AI provider. Knowledge search and document-text extraction also use AI services.

Private-system access and customer-resident storage do not keep AI processing inside your network. The Privacy Policy identifies the hosted AI provider and sets out model-training commitments and configuration exceptions.

Read the Privacy Policy

Retention and deletion

Review retention, deletion, and backup terms before connecting sensitive information. The Privacy Policy and your applicable written agreement describe those commitments.

Automatic cleanup of completed investigations is off by default. Confirm the retention setting and backup policy for your organization before connecting sensitive data.

Read the Privacy Policy

Private access and storage.
Two different choices.

Connecting a private system does not mean all data stays there. Choosing a storage home does not move AI processing there.

Reach private systems.

For supported integrations, an in-network agent connects to your private systems and communicates with Vedin over outbound HTTPS. It does not listen for inbound requests.

  • You still control the agent’s network access and the credentials it uses.
  • Tool results can return to Vedin for processing and recording.
  • Installing the agent does not self-host the Vedin service or its AI models.

Store knowledge in your SharePoint site.

Eligible organizations can set up customer-resident knowledge storage in a shared Microsoft 365 SharePoint site. This is a setup-time choice, separate from connecting a document source.

  • Setup requires administrator approval and completed encryption rollout, before you add the affected content to Vedin. There is no backfill migration for existing content.
  • This stores knowledge content, investigation memory, and code digests, not investigation transcripts or all operational records.
  • Content is still processed by Vedin and the configured AI services. Vedin retains the operational metadata and encryption keys needed to provide the service.
  • If the selected storage is unreachable, resident knowledge is unavailable. There is no silent fallback to Vedin-managed storage.

Confirm storage eligibility and setup before importing content. Customer-managed encryption keys and full self-hosting are not offered by the storage option described here.

Bring your questions.
Review the details.

Tell us what your security or procurement team needs to assess. Confirm requirements before connecting sensitive information.

Privacy PolicyHow information is collected, processed, shared, and retained.Terms of ServiceUse of the service, responsibilities, and applicable agreements.

Security review

Ask about the controls, configuration, and evidence relevant to your organization.

Data processing agreement

Request a DPA and discuss the processing terms your organization needs.

Subprocessor information

Request the current provider list, processing purposes, and location details.

For contact details, see the Privacy Policy.

We’ll discuss your requirements and the terms available for your organization.

Product screenshot

Choose original size to read the details. Scroll the image to explore it.