For IT teams
Say yes to AI without losing control
Every team wants an assistant. You are the one accountable for what those assistants read and repeat. Put one governed source behind them: approval before serving, scoped access per connection, and an audit trail throughout.
- Scoped key per assistant, revocable any time
- Nothing unreviewed is ever served
- Audit trail on every change
- Hosted MCP + REST, nothing to run
Asked
How do I request VPN access?
The approved answer
Submit the Network Access form in the IT portal. Standard requests are approved within one business day; contractor requests need a sponsor.
Asked
Which AI tools are approved for customer data?
The approved answer
Only tools on the approved-vendor list with a signed DPA. Pasting customer data into unapproved tools is a policy violation.
Asked
What's the password policy?
The approved answer
Passwords require 14+ characters, and any shared account needs a manager-approved exception. SSO with MFA is mandatory for all core systems.
One connection pattern for every team
Each assistant gets its own scoped key over hosted MCP or REST. One config block to connect, one place to revoke.
Access enforced at the source
The service checks workspace, key scope, and approval status before returning anything. You do not rely on a prompt to enforce policy.
Your own policies, served too
VPN steps, security policies, and approved-software lists become pages the internal assistant answers from, instead of ticket queues.
Nothing new to operate
The server is hosted. No vector database, retrieval pipeline, or index to run, patch, or scale.
Questions
- What do we have to deploy?
- Nothing. It is a hosted service. Assistants connect with a config block and a scoped key; there is also a REST API.
- How is access controlled?
- Keys are scoped to one workspace and its allowed classifications, support read or write scope, and can be rotated or revoked at any time.
- Can we audit what changed and who approved it?
- Yes. Every page keeps author, reviewer, and approval time per version, and approvals and rejections land in the audit log.
- Where do the deeper technical docs live?
- The MCP page and the docs cover the endpoint, tools, authentication, and rate limits in detail.
Planning AI rollout across many teams?
Tell us about your environment and security review, and we will help you plan it.
Put a governed source behind company AI
Start free, connect one assistant, and bring your security team when you are ready.
Connect in a minute
Create a token from Settings → API Tokens, then point your MCP client at the hosted endpoint.
{
"mcpServers": {
"curated-data": {
"type": "http",
"url": "https://curateddata.megacorp.company/api/curated-data/mcp",
"headers": { "Authorization": "Bearer ckd_live_••••••••" }
}
}
}