An internal brand-compliance and AI-assisted writing platform that keeps a global brand's voice consistent across every communication, with no external API.
Across an organization operating in 14 countries, hundreds of people write content on behalf of the brand — press releases, presentations, product sheets. Manually checking every text against the Brand Book doesn't scale, and routing that writing through external AI tools raises data confidentiality and cost concerns an industrial company can't ignore.
Instead of wiring the platform to an external LLM, I built a rule engine that translates the company's Brand Book — tone of voice, terminology, capitalization, spelling, inclusive language, CTAs — into an audit-and-rewrite system that runs entirely in the browser, without sending any text outside the organization.
The result is the AI Brand Platform: a Writing Assistant, Tone Auditor, spelling corrector, a product knowledge agent, and a technical cable decoder. Anyone in the organization can check and improve a piece of writing in seconds, without depending on the Design team or on external tools.
Rewrites sentences according to Brand Book rules — tone of voice, spelling, capitalization and brand terminology — with instant feedback inline.
Analyzes text sentence by sentence and flags tone-of-voice deviations before publication, with a clear report of what to fix.
An AI agent that answers natural-language questions about the product portfolio, backed by a structured technical database.
Translates technical codes across international standards (DIN, JASO, SAE, ISO), from a database generated out of internal backoffice files.