It is an adoption fix, not a content one
Most teams respond to unused documentation by writing more of it. The page that would have helped usually exists; what is missing is it arriving in time.
Resolution Agent
Confluence, at the moment it matters
Your best documentation is probably already written and sitting in Confluence, where nobody looks during a live ticket. This is Ascendo reading across your spaces and putting the relevant passage in front of an agent while they work, without anyone leaving the case to go and search for it.
- Reads across your spaces, not one page at a time
- The relevant passage arrives during the ticket
- Nobody leaves the case to go searching
- An adoption fix, not another content project
The documentation exists. Nobody reads it during a ticket.
Confluence is where most technical organisations keep their best material — runbooks, product documentation, the page somebody wrote after a bad incident. It goes almost entirely unread during a live ticket.
That is not laziness. Searching it while a customer waits means leaving the ticket, guessing at the author’s vocabulary and reading several pages to find which applies. So the organisation concludes it has a documentation problem. It usually has a retrieval problem, and writing more pages makes it worse.
Bringing the page to the ticket
The material is already written. All four of these are about it arriving in time.
- Step 1
Index the spaces
The spaces you nominate are read in and kept current, so retrieval reflects the page as it is now.
- Step 2
Search on the problem, not the wording
Retrieval runs on what the ticket is about, so the right passage is found without guessing which words the author chose.
- Step 3
Read across spaces at once
Answers are frequently split — the procedure in one space, the caveat in another. Both are found together.
- Step 4
Deliver it into the case
The passage appears where the agent is already working, which is the difference between documentation being available and being used.
Why this is not a documentation project
The page you needed almost always already existed.
Confluence is one source, not the only one
Restricting an answer to Confluence would rebuild the silo. The same request also reaches your resolved tickets, your manuals and whatever else you have connected.
The page stays the source of truth
Nothing is copied into a second system to go stale. Answers cite the page they came from, so the author keeps ownership.
Why the usual approaches fall short
Three of these lose to asking a colleague, which takes four seconds.
Confluence search itself
Matches the author’s words against the agent’s guess, and returns pages rather than passages.
Copying content into a support tool
Creates a second copy that diverges. Within a quarter nobody can say which is current.
Asking a colleague
Fast, accurate, and it consumes two people. It is how the same question gets answered forty times.
Writing more documentation
The reflex response, and it addresses authoring rather than retrieval.
What things are called
Confluence vocabulary, and the two Ascendo terms it meets.
- Space
- A Confluence area holding related pages, and the unit you nominate for indexing.
- Passage retrieval
- Returning the relevant part of a page rather than the page itself.
- Data source
- A connected system Ascendo indexes and reasons over. Confluence is one among several.
- Citation
- The link back to the source page, so an answer can be checked and the original opened.
- Mapping
- Associating a source with products or product groups, which also controls who can see it.
Frequently Asked Questions
No, and you should not. Confluence stays the system of record and the author keeps ownership. Ascendo indexes it and cites back to it, so there is no second copy drifting out of date behind your back.
Resolution Agent
Test it on your own spaces
Point us at the Confluence spaces your team relies on and we will show you what Ascendo answers from them — including the pages nobody has opened in a year.
Talk to us