Measured Launches Facet: AI-Assisted Procedural Brand Image System Built with Agentic Coding
Key Takeaways
- ▸Facet uses procedural generation (not AI image synthesis) to create visually consistent brand images constrained by the Measured visual system's core shape grammar
- ▸The project demonstrates effective use of agentic engineering—where AI coding agents amplify expert developers' work—with agents writing p5.js code iteratively while humans maintain visual and quality oversight
- ▸By combining agent-written code with in-browser review and human-controlled design constraints, the team achieved professional-grade output without sacrificing brand consistency or UI standards
Summary
Measured has introduced Facet, a procedural brand image generation system built using agentic coding workflows rather than traditional AI image generation. The tool was developed to extend the Measured visual brand system—a flexible framework based on core geometric shapes—by automatically generating new images for blog posts and campaigns while maintaining strict brand consistency. The project represents a novel application of agentic engineering, where AI coding assistants wrote the bulk of the p5.js generation code without traditional code review, while design engineers retained control over UI/UX and visual standards. By combining procedural generative art principles with AI-assisted development, the team was able to bridge the steep learning curve of Processing-style development while maintaining the high design standards expected of production work.
- This approach bridges the gap between complex generative art systems and practical business needs, allowing design engineers to produce outputs in domains (Processing/p5.js) where they previously lacked expertise
Editorial Opinion
Facet represents a thoughtful middle ground in AI-assisted development: neither fully automated nor traditionally hand-coded. By letting agents handle code generation while maintaining human judgment over visual outcomes and brand constraints, Measured demonstrates that agentic engineering excels not when replacing human expertise, but when amplifying it. This model—bounded by clear design principles and subject to in-browser iteration—could become a template for how creative and design-focused companies leverage AI coding tools without compromising quality or brand integrity.
