How it works
From the manual you have today to a reviewed topic, a video and versions that machines can read. People approve every step that matters.
In development The platform is in development. This page describes its design.
Five steps
- Step 01 / 05
Import
Bring in manuals as DITA/XML, PDF or Word. The platform proposes topics and steps from PDF and Word files.
- Step 02 / 05
Edit and review
Your editors change the proposed steps and translations. A reviewer who did not edit the topic approves it, or two different reviewers if you require that.
A person reviews and approves
- Step 03 / 05
Generate
The platform builds the web page, Markdown, structured data, video script, subtitles, audio and video from each approved topic.
- Step 04 / 05
Deliver
You decide per manual or per topic what is public. Public pages are served from your own deployment, on your domain.
A person reviews and approves
- Step 05 / 05
Measure, if you choose
See which bots read which public topics, and check how assistants answer real questions. Results show trends, not guarantees.

One topic, every output
When a reviewer publishes a topic, the platform is designed to generate each of these and keep them in step with every change.
One approved topic
- Web page: the official answer on your domain, readable without JavaScript.
- Structured data: tells machines what the page is: an article, its steps, the video, the product and the publisher.
- Markdown version: the same text, clean, at the page address plus .md.
- Video: shows the physical steps, on the page and, if you choose, on your own video channel.
- Subtitles and transcript: make the video readable for people and machines.
- Audio: a read-aloud version for hands-free use and accessibility.
- Sitemap, robots.txt and llms.txt: help crawlers find every public page.
- MCP: lets assistants ask for the current version of a public topic directly.
Five principles
One topic, one question, one URL
Each task, concept, error code or question for a specific model gets its own stable web page.
The answer comes first
The first two or three sentences answer the question completely and name the brand and the model.
One source of truth
Each topic exists once as structured content. The web page, Markdown, video, subtitles, audio and MCP response are made from it, so they do not contradict each other.
One official page
Every topic has one canonical address, on your own domain or on your subdomain of the platform domain.
Text first, media as support
Assistants read text. Video and audio help people, and count for assistants through the text around them: titles, transcripts, subtitles and chapters.
Rules for review
- Anything the platform proposes is labelled AI draft until a person approves it.
- A reviewer who did not edit a topic approves it. You can require two different reviewers for topic text; shared warnings and components always need two. The database enforces it, not only the screen.
- Warnings, cautions, torque values and part numbers are never reworded by automation without a person's approval.
- When a new revision of the source arrives, the platform shows what changed and marks the affected steps Needs update.
