Most AI content failures aren't model failures. They're system failures the model inherited.
This piece introduces SignalScale: a diagnostic framework for assessing how content systems behave under AI interpretation. Drawing on enterprise case studies including NYC MyCity and Air Canada, it reframes content quality as structural evidence, not editorial opinion.
Published through Content Science in collaboration with founder Colleen Jones.
Designed and shipped a fully functional web app that turns allowance into a real economy for kids. Built multi-kid switching, parent/kid modes, and behavioral loops that make money feel tangible. Validated through 6+ months of daily use with a real child.
Led content strategy for the CVS mobile homepage serving 60M+ members, building an omnichannel content system that unified pharmacy regulation, retail conversion, and healthcare empathy.
Created frameworks that flex across clinical accuracy, promotional urgency, and trust-building without breaking tone. Proving that emotionally intelligent content scales in regulated environments.
Led Writer AI implementation for a 60M+ member healthcare platform, building the prompt libraries, content infrastructure, and governance frameworks that made enterprise AI usable in a regulated environment. Then drove adoption across clinical, legal, compliance, and product teams.
AI doesn't fix broken systems. It accelerates them.
SignalScale™ is a diagnostic framework that reveals whether your content infrastructure is ready for AI implementation, or if AI will amplify existing structural problems at scale.
Built while collaborating with LLM and data science teams at CVS Health, the framework scores organizational readiness across Structural, Strategic, Operational, and Ethical and Inclusive dimensions.
The core insight: If your system can't reliably serve a user with a disability, it can't reliably serve AI. Accessibility isn't a compliance checkbox. It's the diagnostic that reveals every structural failure.
Content reveals organizational truth.
Bad labels aren't writing problems. They're evidence of unclear ownership. Inconsistent tone isn't a style issue. It's proof of misaligned teams. Vague error messages aren't poor UX. They're symptoms of technical debt no one wants to name.
Forensic UX is a methodology for reading content as diagnostic evidence. It teaches you to see patterns in language that expose structural problems, power dynamics, and decision-making failures.
Because fixing the content without fixing the system just creates better-written dysfunction.
Published in Bootcamp, June 2025
This article argues that without strategic alignment across systems, governance, and intent, AI doesn’t create clarity — it amplifies chaos. It reframes content strategy as the primary lever for trust, coherence, and scale in an AI-saturated environment.
Published in Ditto – October 2025
This article challenges efficiency-first narratives around AI and makes the case for content designers as strategic partners in shaping how AI systems communicate, reason, and behave. It outlines a more responsible, design-led approach to AI adoption in real product environments.
Published in Bootcamp – July 2025
This piece reframes accessibility as a structural readiness test, showing how AI magnifies invisible design debt, behavioral blind spots, and governance gaps across complex content systems.
Published in Bootcamp – October 2025
This article introduces a diagnostic framework for assessing AI readiness across content teams. It shows how AI implementation exposes invisible design debt, behavioral blind spots, and governance gaps across complex content systems — and what to fix first.
Whether you're building AI content systems, untangling organizational complexity, or just want to talk about the state of content design, I'm up for a conversation.
Send me a message below, or email me directly at iantrichards@icloud.com
Connect on Linkedin