How do you preserve technical correctness when information is changing, distributed, and AI-assisted?
Context
Technical documentation becomes part of an engineering information system when other people rely on it to make decisions. My work at Dynamo Air involved drafting and revising material in a Git-based Markdown repository, using AI assistance while keeping engineering review central.
What I did
- 01
Worked with repository-based technical documents rather than isolated generated drafts.
- 02
Used Claude Code, Markdown, branches, commits, and reviewable changes to develop and revise material.
- 03
Reviewed consistency across assumptions and references, then incorporated engineering feedback.
Architecture / workflow
A draft is a proposal. Review gives it context.
Who judges technical correctness?
- 01INFORMATION / ARTIFACT
Technical sources
Documents and engineering feedback
- 02MODEL ASSISTANCE
Assist the draft
AI-assisted writing and consistency review
- 03SOFTWARE LOGIC
Expose the change
Markdown, Git branches and reviewable diffs
- 04HUMAN JUDGMENT
Engineer review
People retain technical authority
INSPECT THE FAILURE BOUNDARYA plausible draft contains a stale reference or incompatible assumption.
Review must examine the changed claim and its source. Version history alone cannot establish correctness.
An abstraction of the documentation workflow. No aircraft design, certification authority, or proprietary document is depicted.
Where it gets difficult
A fluent draft can conceal incompatible assumptions. The useful unit of review is the changed claim and its source, not how convincing a paragraph sounds.
Source control makes differences visible, but it does not establish technical truth. Review needs both a traceable change and the right technical reviewer.
Limits & evidence
- I am describing documentation and review work, not aircraft design authority or certification responsibility.
- Company engineering documents and technical specifications are withheld. The public account describes the process, not the aircraft.
- Repository activity confirms participation; it does not establish every employment date or the downstream effect of a particular correction.
Source-controlled engineering documentation
Local repository and authored commit history support participation in Git-based technical documentation. The supplied account and prior portfolio review describe AI assistance and engineering review.
What this does not establish
Company documents remain private. Repository participation does not establish design authority, certification responsibility, exact employment dates, or measured outcomes.
What this changed
AI can accelerate the production of technical material. The harder work is keeping assumptions explicit and making every meaningful change reviewable by someone with the authority to judge it.