A product and AI strategy studio built on behavior design, obsessive execution, and twenty years of making things people actually want to use.
From workflow modernization to AI strategy. Startups to enterprise.
Embedded leadership that turns ambiguous problems into shipped outcomes. Not a deck and a handoff.
Twenty years of UX applied to AI orchestration. We design the workflows humans actually trust.
User feedback drives every decision. Test with real people before you commit to a roadmap or burn more budget.
Engagements range from focused sprints to embedded fractional leadership.
Happy to discuss our approach directly.
Built on twenty years of product leadership at Apple, CVS Health, Disney, Take Two, and two startups from zero to shipped.
Local clinicians know their patients better than anyone, but they're stretched impossibly thin. Digital solutions have failed to help, be it for triggering shame or being culturally irrelevant.
MotiSpark scales the power of local provider by sending timely personalized videos to patients when and where they need them along with delightful personalized content.
Partnered with Cedars-Sinai, Merck, and UCLA. Deployed across 7 states.
The result: telemedicine appointments increased 4x.
View the patentLed product and design across Retail Pharmacy, Specialty, and Caremark to unify fragmented member identity into a single CVS Health ID. Guided 8 scrum teams across 3 lines of business, reaching 47M+ members.
Publicly shareable metrics only.
Most client work lives under strict NDA. These are side projects we built to solve our own real-world problems. Happy to share our solutions.
Modular steel-and-living-biomass boundary that replaces combustible fences with passive fire and flood defense. Geothermal cooling, harvested rainwater, zero power. Patent pending.
Born from real need after a family loss. Vision models for identification, Python for pricing math, an LLM only where language generation matters. Concept to live product in under 12 hours.
Every client engagement teaches us something about how AI systems succeed or fail at representing human intent. That applied knowledge feeds directly into our work on the standards and infrastructure being built right now.
AI agents are beginning to act on our behalf: shopping, booking, transacting. Google, Shopify, OpenAI, Stripe, Anthropic, and Mastercard have all deployed agent commerce protocols in the last year. What hasn't shipped: the user intent layer. No standard exists for how preferences are structured, verified, or controlled as they move between AI systems and the platforms they interact with.
A verified agent executing corrupted preferences is a trusted system acting against the user it serves. We're contributing to the standards that will close this gap.
The right to own and audit your algorithmic self so “biases” serve your intent, not a platform’s profit.
Agency over Extraction: Shifts the AI from a tool of surveillance to a tool of self-actualization.
Auditability: Converts “black box” algorithms into transparent, user-steerable models.
Provides the technical mediation layer required for secure preference exchange across disparate agentic protocols like the Model Context Protocol (MCP), User Context Protocol (UCP), and Agent Communication Protocol (ACP).
Grants individuals the “Master Key” to their digital twin, ensuring AI biases align with personal intent rather than platform profit while providing more secure and efficient agents.
Formal response to the AI Agent Standards Initiative, framing user intent integrity as foundational to both agent identity security and efficient agent workflows.
Read the full response
Whether it's modernizing a platform, bringing a wild idea to something tangible, fixing a churn problem, or trying to grasp what our future holds, we're here.