CHALLENGE:
Most businesses exploring AI for marketing and creative work quickly run into the same wall: without proper context, AI outputs require constant re-prompting, corrections, and oversight before they're usable. There was no scalable way to ensure outputs reflected a brand's voice, channel requirements, compliance rules, or tone — and the burden of figuring all of that out fell entirely on the individual user, every single time.With 30 years as a trusted name in plant-based meals, Dr. Praeger's had recently updated their brand identity and packaging. Their digital presence needed to reflect the new direction while solving a core business problem: their products are sold in stores, not online, so the website needed to drive customers to retail locations.
APPROACH:
As Partial Owner, Product Design & Guardrail Architect, I recognized Synthia's core technology — a rules-based system that sits between the user and the LLM — as a powerful but underdeveloped platform for marketing and brand creative. The system needed substance: a library of reusable, ready-to-deploy guardrails that new users could put to work immediately, without deep configuration or technical setup. Working alongside my partner Alex at AXM, we built a library of nearly 400 guardrails across categories including Channel, Tone, Brand, Benefit-Driven, Social, and Analytics. We then introduced "presets" — curated groupings of guardrails designed to solve entire workflows out of the box. A CPG brand drafting a press release, for example, can combine a brand guardrail with press release formatting and FDA compliance rules in a single preset, producing on-brand, compliant drafts from the start. Beyond content architecture, my design fingerprints are woven throughout the current product — from UI refinements and admin functionality to client setup flows and overall organization — all shaped around one principle: deceptively simple, incredibly powerful.
RESULT:
A platform that eliminates the cycle of re-prompting by giving AI the context it needs to get closer to final on the first pass. New users onboard into a robust, immediately useful system. Clients with specific needs get precision-built guardrail stacks that scale with them. What was once an promising but bare framework is now a scalable AI infrastructure for marketing teams that want consistent, brand-right output — without making every employee a prompt engineer.