Human-AI Systems Design

Intelligence is
no longer scarce.Understanding is.

I help organizations understand what actually changes when AI becomes part of how people work — not just the capability side, but the human side. The decisions, the habits, the trust. The stuff that determines whether any of it sticks.

Current Global Head of Design, H-AI & Emerging Systems
Base New York
Open to Principal / Director — Frontier Labs & AI-native organizations
Scroll
00 — Position

Most organizations don't have
an AI problem.
They have an understanding
problem.

At ?What If! / Accenture, I lead teams that work on a deceptively hard problem: how do you make AI genuinely useful inside real organizations, with real people, under real pressure?

That means designing the workflows, the moments of handoff, the narratives that let people see themselves in a future before it's built. It means helping leadership teams ask better questions — not just "what can we automate?" but "what should stay human, and what happens at the boundary?"

I've done this across retail, healthcare, aviation, financial services, and manufacturing. The industries differ. The core challenge doesn't: people need to recognize themselves inside the future before they can trust it.

My work: making the future legible to the people who have to live inside it.
01 — Experience
2019 — Present
Accenture / ?What If!
Global Head of Design, Human-Centered AI & Emerging Systems
I lead teams working at the intersection of AI capability and organizational reality — figuring out not just what's technically possible, but what people will actually use, trust, and build their work around. Most of my time is spent with C-suite leaders, helping them see what intelligent systems mean for the people in their organizations.
Agentic AI Systems design C-suite engagement Anthropic API Interactive prototyping
2014 — 2019
?What If! Innovation Partners
Creative Director
Before AI became the central question, I spent years learning how people actually think and behave inside organizations — across retail, financial services, healthcare, consumer goods, and public sector. That foundation shapes everything I do now. The technology changes. Human nature doesn't.
Narrative prototyping Interaction futures Adoption architecture
02 — Evidence (six cases)
001
Starbucks: Designing AI That Protects the Human Moment
Seventy percent of U.S. Starbucks transactions happen in the drive-through. The challenge was not how AI could accelerate the interaction — it was how AI could protect the conditions that make connection possible. We designed agentic support systems that surfaced signals like tone shifts, timing pressure, and loyalty context not as instructions, but as subtle prompts that preserved the barista's authorship of the interaction.
Signal: Human-AI interaction under real-time pressure
Retail / customer experience
002
Walmart: Agent-Orchestrated Decision Environments for Volatile Supply Systems
Fresh supply chains are shaped by weather variability, logistics timing, demand elasticity, shrink risk, and substitution behavior. Human operators manage these forces intuitively — but rarely see them as a connected system. We designed a multi-agent coordination environment where forecasting, logistics, pricing, and substitution agents produced shared situational awareness. Teams stopped reacting locally. They began responding systemically.
Signal: Agent orchestration across distributed systems
Retail / supply chain
003
Walmart Home Office: Designing the Moment Agents Become Infrastructure
Early AI deployments require explicit interaction. Mature deployments become ambient coordination layers that support decision-making without demanding attention. We designed a behavioral walkthrough showing what work looks like once agents become normal — the moment employees stop asking "what can this system do?" and begin asking "what does the system already know that helps me decide?"
Signal: Behavioral activation of agent ecosystems
Enterprise / workforce transformation
004
Southwest Airlines: Designing the Adoption Layer for Enterprise Intelligence Systems
Organizations rarely resist intelligence infrastructure itself. They resist invisible changes to responsibility. We designed narrative prototypes following a single traveler across an end-to-end journey — not presenting the platform as architecture, but as behavior. The platform stopped being perceived as a marketing upgrade and started being understood as an organizational memory layer.
Signal: Enterprise intelligence as shared memory infrastructure
Aviation / customer intelligence
005
Carhartt: Designing Workforce Modernization Without Breaking Identity
Organizations adopt intelligent systems more successfully when employees see those systems strengthening the continuity of their role rather than redefining it. We designed narrative prototypes showing modernization appearing inside familiar work environments — production teams seeing upstream context earlier, designers understanding downstream constraints sooner, experience accumulated on the floor becoming organizational memory.
Signal: Identity-safe workforce modernization
Manufacturing / workforce transformation
006
Mastercard: Designing Trust Infrastructure for Agentic Commerce Environments
As intelligent assistants begin recommending purchases, managing subscriptions, and coordinating transactions on behalf of people, trust must become programmable. We designed scenario walkthroughs showing how delegation becomes understandable — not how automation becomes invisible. The payment network evolves from transaction processor to orchestration layer between identity, intent, authorization, and action.
Signal: Trust-preserving agentic commerce delegation
Financial services / agentic commerce
03 — Contact

Let's talk about
what comes next.

I'm open to conversations at frontier labs — particularly roles where the question isn't just what AI can do, but how it actually lands with people.

Email
[email protected]
LinkedIn
linkedin.com/in/dawkins-jeremy
Resume / CV
PDF available on request