Curated Data vs enterprise search
Indexing everything is not the same as approving anything
Enterprise search tools connect your files and index them as they are, including the stale draft and the wrong number. Permission-aware retrieval decides who can see a document, not whether it was ever approved. Curated Data adds the step search tools skip.
- Approval gate, not just permission filtering
- Serves only the current approved revision
- Built for AI agents over MCP and REST, not just a search box
- Free tier to start
Permission-aware is not approval-aware
A search tool's access control decides who is allowed to see a document, not whether the document is right. A stale policy you have permission to see is still stale. Curated Data's approval gate is a separate, additional check: only a reviewed and approved revision is ever served.
Search finds candidates. This serves an answer.
A search tool hands back a ranked list of documents for a person to read. Curated Data serves an agent the current approved concept directly, over MCP or REST, at answer time, with no ranking judgment call left to the agent.
Nothing reaches an agent unreviewed
Enterprise search indexes a file the moment it lands in a connected source, as it is. A concept doc in Curated Data stays a draft until a reviewer approves it, so what an agent can retrieve is exactly what your team has vetted.
Built for the AI agent, not retrofitted onto a search box
The hosted MCP server and REST API are the primary interface, with per-token classification and rate limits designed for a machine caller from the start, not added on top of a human search product.
Questions
- We already have an enterprise search tool connected to everything. Isn't that enough?
- It depends what you are asking it to do. If the goal is a person finding a document, that is what search tools are built for. If the goal is an AI agent answering a high-consequence question correctly, a search tool has no way to tell your agent that the top result is a policy version from two quarters ago.
- Doesn't permission filtering already keep bad content out?
- Permission filtering keeps the wrong audience out. It does not keep a wrong or outdated answer out for the right audience. Those are different problems, and Curated Data solves the second one with an explicit approval step.
- Can Curated Data index our existing file sources the way search tools do?
- Curated Data supports bulk import of existing markdown so you are not starting from a blank page. Imported material still enters the review workflow; importing a document does not make it agent-ready on its own.
- Do we need to replace our search tool?
- No. Keep it for people browsing your intranet. Curated Data is the separate, smaller, reviewed source you point your AI agents at for the answers that need to be right.
- What does this cost?
- Free for up to 200 approved pages and 5 members. Team is $99 a month for larger rollouts, and Enterprise is custom priced.
Comparing knowledge sources for an AI rollout?
Tell us what you have connected today and we will help you figure out where an approval gate actually helps.
Search finds it. Approval decides your agent can trust it.
Start free and put your first approved answer behind an agent.
