Work Area
Knowledge and Content Systems
We build knowledge and content systems that keep what your organization knows organized, searchable, and current. Staff find the right answer quickly, and AI assistants can use the same source safely, with every answer traceable to its document.
- 1,200+ clients helped
- Building since 2010
- 15 services, one team
- 50+ technologies
At a glance
What we build
We organize what your organization knows, such as policies, guides, and articles, so people and AI assistants can find the right answer quickly.
Problems we solve
Knowledge is scattered across documents, chats, and people's heads, so answers are hard to find and are often out of date.
Who we serve
Teams that answer the same questions repeatedly, and organizations publishing a lot of content or guidance.
How to choose
Organize around the questions people ask, name an owner for each topic, and set a review date for every page.
In practice
Answers that stay findable and true
We organize the answers your organization needs so they can be found, are owned by named people, and are reviewed regularly so they stay correct.
Most organizations have the knowledge. They lack the structure. Answers sit in old emails and in files named final-v3, and the person who knows retires or moves on.
A good system also prepares your content for AI assistants. They answer better, and more safely, when drawing from clear, current, well-labeled pages than from a heap of documents.
The layers of a knowledge system
Layers from top to bottom. Questions people ask; Search and navigation; Organized content; Sources.
- Questions people askThe starting point, in their own words.
- Search and navigationFinding the answer fast.
- Organized contentPages with one topic, one owner, one review date.
- SourcesDocuments, policies, and expert knowledge.
Often built with
How it is organized
How knowledge is organized
Each page has one job, one owner, and one review date.
How knowledge is organized
A sequence of 5 steps. Collect: Gather what exists, and find what is missing.; Organize: Group by the questions people ask.; Write: One topic per page, in plain words.; Publish: With search, labels, and links.; Review: Owners check each page on schedule..
- CollectGather what exists, and find what is missing.
- OrganizeGroup by the questions people ask.
- WriteOne topic per page, in plain words.
- PublishWith search, labels, and links.
- ReviewOwners check each page on schedule.
How we work
How a knowledge project runs
Find the top questions
We listen to support, sales, and staff to learn what people ask most.
What it produces
- Question list
Design the structure
We organize topics and agree who owns each.
What it produces
- Structure
- Owners
Build the system
We set up the pages, search, and review reminders.
What it produces
- Working system
- Review schedule
Hand over and keep fresh
We train owners and measure which answers are used and which are missing.
What it produces
- Training
- Usage reports
With AI
How AI assistants use a knowledge system
| Use | What happens | What keeps it accurate |
|---|---|---|
| Answering staff questions | The assistant finds relevant pages and summarizes them. | It shows its sources, and staff can check. |
| Helping customers | The assistant answers common questions from public pages. | It is limited to approved content, and hands off when unsure. |
| Drafting replies | It drafts from past answers for a person to review. | A person approves every reply. |
A typical example
One current page for every procedure
A typical case is a team whose procedures are spread across shared folders, with several versions of each. New staff ask colleagues, and answers differ.
A knowledge system could hold one current page per procedure, with an owner and a review date, and search that finds it. An assistant could answer questions from those pages and show where the answer came from.
Technology
Under the hood
The detail a technical reviewer will ask about: how content is structured, how search finds it, and how AI assistants use it safely.
Content is stored in a content management system with consistent fields, so it can be shown on a site, searched, and supplied to other tools.
- Full-text search tuned to the words people use.
- Retrieval-augmented generation, which lets an AI assistant draw on your pages and cite them.
- Permissions, so private content is never shown to the wrong person.
Tell us the questions you answer most
List the questions your team or customers ask most often. We will show how to organize the answers so they stay findable and current.