Knowledge Management
AI Knowledge Base Builder
Turns scattered documents into a structured, maintained knowledge base.
Knowledge Management
AI Knowledge Gap Finder
Compares what people ask against what your documentation actually covers, and reports exactly what to write next.
Most teams know their documentation has holes, but not where they are. AI Knowledge Gap Finder compares the questions people really ask, from tickets, search logs and chat transcripts, against what your documents cover. It answers what it can with a citation and flags every question it cannot answer well in a clear gap report, so your writers know exactly what to work on next.
- A gap report showing exactly which questions go unanswered
- Cited answers for everything your documents already cover
- Writers focus on the content customers and staff actually need
$8,900
20–30 working days
Optional maintenance $445/mo
What it is
AI Knowledge Gap Finder is a cited Q&A assistant that also analyzes your support tickets, search logs and chat transcripts against your documents, so every low-confidence or unanswerable question is flagged in a gap report instead of silently guessed at. You get a monthly list of what to write next.
Why businesses buy it
Documentation work is often guesswork: teams write what they assume people need, while the questions that really cost support time go unanswered. Every gap becomes a ticket, a call or a wrong guess. AI Knowledge Gap Finder shows you, from real questions, which topics are missing or outdated, so writing effort goes where it reduces support load most. It also works as a cited Q&A assistant from day one. The Scope Card fixes the sources, document volume, permission groups and testing before you order, so you know the full scope upfront.
Problems this solves
You can't tell which questions your documentation fails to answer
The tool compares support tickets, search logs and chat transcripts against your documents. Questions it cannot answer with confidence are flagged in a gap report instead of guessed at.
Writers choose topics by hunch rather than evidence
Each month you receive a list of what to write next, drawn from real questions people asked. Writing effort can go where the gaps actually are.
A Q&A assistant that guesses can spread wrong answers
Low-confidence questions are flagged rather than answered loosely. Answers it does give are cited, and a 60-question benchmark with an accuracy report checks quality before launch.
A good fit if
You already have a Q&A assistant or documentation, but nobody knows which questions it actually fails to answer well.
You have support tickets, search logs or chat transcripts showing what people really ask, not just what you assumed they would.
You want a structured report of missing or outdated documentation, not just a working chatbot.
You are ready to act on the report, assigning someone to write the documentation the report says is missing.
Not the right fit if
You just want a working Q&A assistant, without the gap analysis on top. See AI Company Knowledge Hub, or AI Company Memory.
You already know what is missing and just need it written. See AI FAQ Generator, which drafts the content directly.
Your documentation does not exist in any structured form yet, so there is nothing to compare questions against. See AI Knowledge Base Builder.
You want a single narrow document set answered, not gap analysis across your whole knowledge base. See AI Document Q&A.
How it works
- We connect up to 3 sources: your documentation, such as Google Drive or Confluence, plus a source of real questions, such as support tickets, search logs or chat transcripts.
- Up to 2,000 documents are indexed for cited Q&A, so the assistant can answer questions the documentation already covers, with a citation on every answer.
- Every real question is checked against that index. A confident, well-supported answer is delivered normally; a low-confidence or unsupported one is logged instead of guessed.
- The logged questions are grouped into a gap report: which topics come up often with no good answer, and which documents look outdated against current questions.
- Up to 3 permission groups keep the underlying answers scoped correctly while the gap report itself stays visible to whoever owns documentation.
- We test the whole setup against 60 benchmark questions with an accuracy report, and a monthly refresh keeps both the index and the gap report current.
What you provide within 72 hours
Your portal lists these the moment you order. Delivery starts the working day after everything is in and your access is checked.
- Access to your documentation sources, such as Google Drive, SharePoint, Notion or Confluence.
- A source of real questions, such as support tickets, search logs or chat transcripts.
- Your user list and the permission groups the underlying answers should respect.
- 30 sample questions, with the answers you expect, to guide the accuracy testing.
- Admin access and API keys for every connected system, shared by adding our team as a user or creating a key you control, never by email.
- Your business information (company basics, hours and brand voice), shared during intake.
- One decision-maker who owns documentation and will receive the gap report.
- An API key for your own AI model account (for example OpenAI or Google Gemini), created for this build with a spending limit.
The 72-hour rule. Send everything on your checklist, including access, within 72 hours of ordering so your build starts on time.
Timeline
Start
Intake confirmed, with documentation sources, question sources and permission groups agreed.
Weeks 1–2
Documentation indexed and real questions ingested from your connected sources.
Week 3
Cited Q&A, permission-group logic and the gap-analysis layer built and connected.
Weeks 3–4
Tested against 60 benchmark questions, with an accuracy report and a first gap report delivered.
Week 4
Your review and two rounds of changes.
Go-live
Live, with a recorded handoff session, and 30 days of support starts.
| Option | Delivery time | Price |
|---|---|---|
| Standard | 20–30 working days | $8,900 |
| Express | 15–25 working days | $10,000 |
| Rush | 10–20 working days | $11,150 |
What you receive at handover
- A cited Q&A assistant in one interface (web app, Slack or Teams) for up to 50 users
- Connections to up to three content sources, such as tickets, search logs and chat transcripts
- A monthly gap report listing what to write next (recommendations; writing the content is not included)
- Test report with 60 benchmark questions and an accuracy report
- Admin guide and recorded handoff session
- 30 days of post-launch support
Expected benefits and how to measure them
- Helps you write documentation people actually need, in the right order.
- Can reduce wrong answers by flagging uncertain questions instead of guessing.
- Gaps closed each month
- Track how many items from each monthly gap report have a published answer by the next report.
- Unanswerable questions
- Compare the count of flagged low-confidence questions in the first gap report with later reports.
- Tickets on undocumented topics
- Tag tickets on topics with no documentation for a month before launch, then repeat a few months later.
Before you start: Export a month of tickets and search logs and note which topics already have documentation. Record how many tickets fall on undocumented topics.
These are things to track, not promised results. Outcomes depend on your data, your team and your market.
Running costs
You pay for the build once. Running Knowledge Gap Finder costs roughly $50 to $300 a month in your own accounts, covering AI model usage, the vector database and hosting, and it moves with how many questions get analyzed each month.
AI usage during the build and testing is billed to your own accounts too, so every provider relationship stays in your name.
Ask about running costsCustom requirements
If you need more than 3 connected sources, more than 2,000 documents, or a second interface, add a note at checkout for a written Change Quote. Your order still runs on its normal timeline. Larger or bespoke projects go through the Custom AI request form.
Request a custom versionKeep it running
We recommend monthly maintenance: monitoring, fixes and monthly change hours, covering a new question source or an updated gap report format. Maintenance starts after the 30 days of support and can be cancelled any month.
$445/mo (5% of the build price a month, includes 4 change hours)
Optional. Starts after the post-launch support period. Cancel any month.
How maintenance worksHow this compares
| Do it yourself with a SaaS tool | Hire a custom agency | Upmark Labs | |
|---|---|---|---|
| Price | A helpdesk analytics tool charges per agent seat every month | A documentation audit from an agency is quoted per engagement | Ours is one fixed price, shown before you buy |
| What it measures | A helpdesk analytics tool reports ticket volume, not documentation coverage | An agency audit depends on what was scoped and how deep it went | Ours compares real questions against your documents and reports the gaps |
| Output | A helpdesk analytics tool gives you a dashboard to interpret yourself | An agency delivers a written report on its own schedule | Ours delivers a working Q&A assistant plus a monthly gap report |
| Delivery | A helpdesk analytics tool is instant to sign up but needs manual analysis | An agency audit usually takes several weeks | Ours has a fixed delivery time shown before you buy, with faster options |
Questions about this product
How is AI Knowledge Gap Finder different from AI Company Knowledge Hub?
AI Company Knowledge Hub is built purely to answer questions well. Knowledge Gap Finder does the same cited Q&A, then adds an analysis layer that compares real questions from tickets, search logs or chat transcripts against your documents, and reports exactly what is missing or outdated. Choose Gap Finder when the goal includes improving documentation, not only answering questions.
What does the AI Knowledge Gap Finder report actually contain?
A grouped list of topics that come up often with no confident answer, plus documents that look outdated against current questions. It is delivered alongside the working Q&A assistant, refreshed monthly, so documentation owners see where to focus writing effort next instead of guessing from memory.
Do we need existing documentation for Knowledge Gap Finder to work?
Yes, some. Knowledge Gap Finder compares real questions against documents you already have, up to 2,000 of them, to find what is missing. If you have no documentation at all yet, start with AI Knowledge Base Builder to create the structure first, then add gap analysis later.
How much does AI Knowledge Gap Finder cost to run each month?
Roughly $50 to $300 a month in your own accounts, covering AI model usage, the vector database and hosting. It moves with how many real questions get analyzed and how many of your up to 50 users ask questions. The build itself is a one-time payment.
Where do the 'real questions' for the gap report come from?
Wherever your team already records them: support tickets, search logs from your help widget, or chat transcripts. You choose the source during intake, and we connect it alongside your documentation so every question can be checked against what actually exists.
How long does AI Knowledge Gap Finder take to launch?
Delivery runs on a fixed schedule from a complete intake, and the delivery panel shows that date plus express and rush options before you buy. Most of the time goes into indexing your documentation, connecting your question source, and testing against 60 benchmark questions before your review.
Find out what your documentation is missing, not just what it says.
Order AI Knowledge Gap Finder and it goes live on a fixed delivery date, with a monthly report on what to write next.
Works well with
