Tech • IA • Crypto
DoorDash gave every employee Claude Code
7/10TL;DR
DoorDash is rapidly expanding AI-assisted coding across its workforce, reporting major gains in development speed while rethinking processes, team structures, and software delivery.
KEY POINTS
Company-wide AI adoption accelerates productivity
DoorDash has rolled out AI coding tools broadly to raise baseline technical fluency across the organization, including non-engineers. Internal adoption has been linked to “massive increases in throughput,” with employees integrating AI into tools like Gmail, Slack, and calendars to improve efficiency. Leadership emphasized that many workers initially viewed AI as simple chat, but adoption surged once practical workflows were demonstrated.
Inflection point in AI coding capabilities
A significant shift occurred between 2025 and early 2026 as newer AI models became सक्षम of handling complex development tasks autonomously. Earlier attempts to ship production code using AI often failed due to setup limitations, but recent tools now allow engineers to successfully deploy features across multiple programming languages. This leap has made AI-driven coding viable at scale.
Engineers shift from writing code to supervising agents
Developers increasingly rely on AI agents to generate code, with some setting goals to avoid manual coding entirely. Teams often run multiple parallel AI sessions across repositories, using tools like worktrees to prevent conflicts. This shift has introduced new workflows that are faster but more cognitively demanding due to constant context switching.
Smaller teams deliver outsized results
DoorDash found that compact teams of three to five people move significantly faster with AI. In one case, a migration project previously estimated to require four engineers over a quarter was completed by a single engineer in just three weeks. Other product timelines have been reduced by more than half, demonstrating the leverage AI provides.
Cross-functional roles increasingly blur
Designers and product managers are now contributing directly to production code, supported by AI tools. This integration allows teams to move faster by reducing handoffs, with designers refining interfaces after engineers deliver functional implementations. The company is also encouraging engineers to become more full-stack, reducing reliance on narrow specialization.
Processes and bottlenecks are being reengineered
Faster coding has exposed inefficiencies in CI/CD pipelines, code reviews, and organizational processes like product and design reviews. DoorDash is investing in AI-powered code review agents and rethinking approval workflows to keep pace with increased output. Security concerns have also grown, prompting parallel investment in AI systems to detect vulnerabilities before deployment.
Internal platform enables secure AI workflows
The company built an internal system called Flux, which provides secure, cloud-based environments for running AI coding sessions. These environments are pre-approved by security teams and power internal tools such as automated code review agents, ensuring compliance while maintaining speed.
Cultural shift toward experimentation and learning
Leadership encourages employees to discard outdated approaches and continuously re-evaluate workflows as AI capabilities evolve. Teams are urged to document both successes and failures in written artifacts, which can be shared across the organization and reused by AI systems. Identifying “AI champions” within departments has helped drive grassroots adoption.
ROI measured through speed and impact
For engineering, success is tracked through code throughput and faster delivery timelines. However, the ultimate metric is how quickly customer-facing features reach production. In non-engineering roles, ROI remains exploratory, with focus on both incremental productivity gains and targeted automation of high-impact workflows.
Leadership emphasizes hands-on engagement
Executives and engineering leaders are encouraged to use AI tools directly to understand their impact and guide teams effectively. This approach helps address resistance and aligns leadership with the rapid transformation of technical workflows.
CONCLUSION
DoorDash’s experience shows that AI-driven development is not just increasing coding speed but forcing a broader transformation of teams, processes, and organizational design to fully realize its benefits.