Niches · Company knowledge & data catalogues
Find a document that was added or changed recently
No company's docs show a full fit yet. The listings below do the job with a limit, or not at all.
What counts as done
A document created or changed at time T is returned by search within a stated lag, including edits, moves and thread replies, and each result carries its last-updated time.
Measured, for our own tests: the lag meets the 15-minute target. No docs we read state these figures, so they don't count against a listing here.
How agents ask for it: the new policy went up this morning, what does it say
, has anyone replied in that thread since
, find the file Sam just moved into the shared folder
Does the job, with a limit 7
It does the job, but something in the docs or a user's account will get in an agent's way. The limit is named in each row.
| Listing | Why |
|---|---|
| Dust B Permutation Labs | Connections sync automatically and docs say an update usually takes a few minutes, and documents carry a timestamp. No per-connector interval or handling of moves and thread replies is published, so the 15-minute target is a claim, not a figure. source |
| Gemini Enterprise B Google LLC | Federation reads sources live, but ingested stores sync incrementally every 3 hours at best, and incremental syncs miss deletions and identity changes, so the 15-minute target holds only for federated sources, with lower search quality. source |
| Glean B Glean Technologies, Inc. | Refresh rates are published per connector (Slack and Confluence event-triggered, Drive incremental every 3h, Confluence 1h) and results carry updateTime, but Glean says events give no time-to-search guarantee, so the 15-minute target isn't shown. source |
| Guru D Guru Technologies, Inc. | Results carry lastModified and cards written in Guru are native, but connected sources sync with only a lastSyncedDate and sync status; no sync interval or lag per connector is published. source |
| Hexis Cloud D Bevelites GmbH | Writes through Hexis land in its clone at once, and POST /api/sync/<branch> from a webhook pulls outside pushes within the request; without the hook a push isn't seen until Hexis touches the branch. Search hits carry no updated time; list_file_changes gives who and when. source |
| Onyx B DanswerAI, Inc. | Refresh Every defaults to 30 minutes, results carry updated_at and accept time_cutoff, and Federated Slack searches live. But in-thread Slack replies (acc-22) and moved SharePoint files (acc-23) are never indexed, silently, and Ingestion API documents don't reach MCP (acc-38). source |
| Lens (Thought Industries) F Thought Industries, Inc. | Answers reflect each source's last sync: manual on Free, weekly on Pro, twice weekly on Pro+, daily on Enterprise. POST /v1/sources/{id}/sync starts a background sync, but results carry no updated time, only the source's lastSyncedAt. source |
Couldn't check 3
The docs were closed to us, or silent. These carry no mark either way.
| Listing | Why |
|---|---|
| Cohere North D Cohere Inc. | Connector sync cadence and result timestamps would be in the closed developer docs; the public pages name connectors but no sync interval. |
| Overclock F Overclock Technologies Limited | The MCP page and tool schemas give no sync interval, and results carry document and effective dates, not a last-updated or indexed time. No account found. source |
| Secoda E Secoda, Inc. | Nothing in the repo on sync cadence or last-updated time, and Secoda's docs were not re-read under its Terms of Use. |
Doesn't do this job 10
Listings in the category that do something else, with the reason.
| Listing | Why |
|---|---|
| Alation C Alation, Inc. | Doesn't aim at recently changed company documents; no tool searches documents by change time. source |
| Atlan C Atlan Pte. Ltd. | Doesn't aim at recently changed wiki, drive or chat content; knowledge files are uploaded by hand or by save_knowledge_content, which indexes them at once, and no connector catches edits, moves or replies. source |
| Basedash MCP B BaseDash Inc. (Delaware, United States) | No document index; Basedash queries live data sources and has no document freshness or last-updated search. source |
| ClicData MCP F ClicData SAS | No document index; the server queries data held in ClicData's warehouse, not documents. source |
| Collibra B Collibra Inc. | No document index or sync of changed documents; the catalogue holds metadata assets, and no freshness or last-updated stamp on search results is documented for agents. source |
| DataHub C Acryl Data, Inc. (d/b/a DataHub) | Doesn't aim at recently changed company documents; metadata comes in through ingestion runs, and documents written through the API or MCP often never appeared in search (acc-26). source |
| Marmot B Marmot Data Ltd | Doesn't aim at recently changed company documents; catalogue metadata arrives through scheduled plugin pipelines, such as every six hours in the docs' examples, with a Run now button. source |
| OpenMetadata C Collate, Inc. | Doesn't aim at recently changed company documents; catalogue metadata arrives through ingestion connectors, and no lag for Context Center files is published. source |
| Precisely AI Studio F Precisely Incorporated | No document index or sync is described; assets operate on records and sample data. source |
| Procore AI Go F Procore Technologies, Inc. | No agent interface, and the support page states no sync lag for the Procore data it answers from. source |
What stops agents finishing
- sync runs every 1, 4 or 24 hours with no on-demand refresh
- incremental sync misses moves and in-thread replies with no error
- new documents written by the agent never reach search
Around it
Before: work-docs.publish. After: company-knowledge.cited-answer.
Where this comes from
Weight 11 of 100 in Company knowledge & data catalogues: 4 of 38 counted accounts (Onyx Slack replies and SharePoint moves never indexed, DataHub documents not appearing, Onyx Ingestion API not seen by MCP), plus Google's connector guide naming 1 to 24 hour sync cycles
The same as data: /api/v1/use-cases/company-knowledge.fresh-document.json, or the MCP tool get_use_case with id company-knowledge.fresh-document. Other jobs in this category: /api/v1/niches?category=company-knowledge.
For companies
Do agents find, use and choose your tools?
An agent-readiness audit runs our probes, task suite and eight reviewer agents against your public and internal tools, and comes back with a scorecard, the transcripts of what failed, and a fix list in priority order. From $2,500, re-run included. We never take payment to move a rank. We do help companies earn one.