- Internal search
Internal knowledge search
Four Wurqurs index the files you already have, answer questions from them with the document cited, and say plainly when the answer is not there.
finding prior work · 90-employee civil engineering consultancy
15–30 minUnder 30 secRead the engagementLibrarian
Answers from your own documents with citations, respects who can see what, and says plainly when the answer is not there.
by hand$4,400/mo
with your Wurqur$285/mo
Precedent Finder
Finds where this was solved before — the similar matter, the comparable job — and says how similar it actually is.
by hand$2,320/mo
with your Wurqur$153/mo
Help Centre Keeper
Keeps self-serve content current, and reports which articles are read and then followed by a ticket anyway.
by hand$2,133/mo
with your Wurqur$158/mo
Bid Librarian
Finds what you answered last time, drafts this time's version, and flags where the true answer has changed.
by hand$2,100/mo
with your Wurqur$148/mo
One or two is a normal build. Your volumes and your rates, in the calculator.
What it costs you now.
The answer already exists, and the quickest route to it is a person who remembers where it went. That person becomes the bottleneck, and their week fills with questions they have answered before.
The same answer, again
A senior engineer writes out an explanation she already wrote for somebody else, because finding the old one takes longer than typing it fresh.
Solved before, solved twice
A detail gets worked out from scratch on a live job while the answer sits in a project folder from an older one.
It leaves when they do
The person who knows where things were filed retires, and the folder names stop meaning anything to anyone.
They work inside what you already run.
Not listed? Tell us what you run.
Questions.
Who can see what it finds?
Each person sees exactly what their existing permissions already allow. Retrieval runs against your own directory, Entra or Okta, so a document someone cannot open in SharePoint stays out of their answers, and every answer cites the documents it came from.
What does a build cost?
$2,500 to $5,000 for a Wurqur built on the systems you already run, paid once.
Do we have to replace Glean or our wiki?
No, and if your corpus can live in somebody else's cloud, buy Glean, Guru or Notion AI instead of us. They are excellent and they work. These four exist for the case those three cannot serve: the documents may not leave the building, which rules out a per-seat cloud index at any price, so the search runs on your own infrastructure against your own permissions. The wiki stays where it is either way, because a Librarian indexes Confluence and Notion rather than replacing them.
What happens to the things it cannot decide?
They go to a person, by a rule you agree before anything is built. A Librarian escalates a question the corpus cannot answer, rather than inventing one. A Precedent Finder says plainly when nothing comparable exists, and hands you the gap in what you have captured.
How long before it is doing real work?
Two Wurqurs took six weeks: four to look at your data, define the load format, work out matching, build the presentation and deploy; two to shadow-run.