Jason Beck

Design leadership · Organisational change · Deliveroo

Ai transformation

Preparing for the next era of product design

AI represents a fundamental shift in how products can be designed, built and shipped. At DoorDash the design org has instigated a rapid transformation embracing new ways of working.

I was part of the design leadership team directly responsible for translating that strategic direction into practical workflows, expectations and capability development.

The question is not “which Ai tools should designers use?”, it’s “how should the role of Product Design change?”

Large design organisations are not usually constrained by access to technology. They are constrained by habits. As LLMs accelerate software creation, our challenge was to help designers move beyond producing specifications and towards becoming builders, experimenters and strategic product partners.

Turning strategy into practice

Three people standing together

Working as a team

Following the DoorDash acquisition, I was part of the Design leadership team responsible for helping the organisation respond to the rapid emergence of Ai. We worked together to establish a shared direction, align leaders around new expectations, and ensure the transformation extended beyond tooling into how Product Design operated day to day.

Stacked layers representing an operating model

Operating model

Strategy alone does not change organisations. I helped translate high-level ambitions into practical ways of working by embedding Ai into everyday design practice through new workflows, team rituals, expectations and leadership behaviours. The focus was on making new ways of working repeatable rather than relying on isolated experimentation.

Checklist representing design pull requests

Design PRs and code contribution

Rather than treating code as the exclusive domain of engineering, designers were encouraged to contribute working interfaces, collaborate through pull requests (via GitHub) and validate ideas in software before traditional implementation, reducing hand-offs and accelerating learning.

Pyramid representing capability development

Capability development

I coached designers and managers as expectations evolved towards building and validating products. This included developing confidence with Ai-assisted development, strengthening technical fluency, encouraging greater product ownership and helping leaders adapt their coaching to support a new generation of builder-designers.

Transformation from traditional design to Ai-enabled building

Progression measured by leverage

As AI reduced the cost of building, testing and learning, the expectations of designers also changed. Success was no longer measured by familiarity with individual tools, but by the leverage a designer could create across a team and organisation.

Six-stage capability progression from personal foundations to organisational leverage

What changed?

Rising bar chart representing a builder mindset

Building is part of design

Designers have moved beyond producing static mock-ups towards creating working product experiences. Validating ideas in software has shortened feedback loops, improved collaboration with engineering and increased confidence through rapid deployment and therefore speedier customer learning.

Branching arrows representing new workflows

AI is part of everyday delivery

AI-assisted development, GitHub collaboration and Design Pull Requests are embedded in day-to-day product development rather than isolated experiments. Designers increasingly work alongside engineers through shared tools and working software.

Two dice representing different expectations

Leadership expectations have evolved

Technical fluency has became a core design capability rather than a specialist interest. Design leaders have coached implementation, experimentation and AI-assisted product development alongside craft and customer thinking.

Rising line chart representing higher leverage

The organisation is more capable

The objective was never simply to make individual designers faster. It was to build an organisation capable of learning, building and experimenting at greater speed and scale, where AI amplified collective capability rather than individual productivity.

Measured over the Q1 and part of Q2 there has been a 60% increase in the rate of experiments being deployed, growing every month.

Many of these are the ‘paper cuts’ usually consigned to the bottom of a product backlog but with designers now able to design and ship these are able to finally address these. Some are brand new features that would have been deemed too costly to devote engineering time on but can now be explored quickly with users to mitigate development costs.

Reflection

AI changes the economics of product design

The biggest shift was not that designers gained new tools. It was that the cost of building, testing and learning collapsed. That changes what product teams can afford to explore, what leaders should expect from designers and ultimately how design organisations create value.

AI made us faster, but it also reinforced the importance of the things it cannot replace: curiosity, judgement, coaching, collaboration and shared understanding. There’s a danger that using AI can promote more solo working when in fact, we need more human connections than ever. Good leaders know how to balance the two.

The organisations that succeed will not be those with the best AI tools. They will be those that redesign the role of Product Design around faster learning.