Building the most AI-pilled engineering team in the world | Fiona Fung (Manager of the Claude Code and Cowork Teams)
Summary
Fiona Fung, Manager of Anthropic's Claude Code and Cowork Teams, highlights the profound transformation in software engineering, noting that Anthropic engineers now ship eight times more code per quarter than in 2021-2025. This shift means coding is no longer the bottleneck, enabling greater ambition and necessitating a focus on verification and quality. Her team leverages Claude for various functions, including management, code review, and automating daily tasks through "routines" to facilitate asynchronous work. Anthropic identifies latent demand, exemplified by the launch of Claude for Small Business, and prioritizes hiring "creative builders with product sense" and "deep systems experts." Fung emphasizes a growth mindset, continuous "dogfooding" of products, and proactive quality frameworks like "bad vs. sad." The team has also moved from six-month roadmaps to "just-in-time" monthly planning. Key challenges include preventing skill atrophy, combating loneliness in AI-native teams, and preserving team culture amidst rapid growth.
Key takeaway
For Engineering Managers navigating AI-driven development, embrace AI tools like Claude routines to automate workflows and achieve significantly higher code output. Focus your team on ambitious problem-solving, proactive quality assurance using frameworks like "bad vs. sad," and continuous "dogfooding" to maintain product pulse. Adapt planning to a "just-in-time" monthly cycle, and actively foster a growth mindset and strong team culture to combat potential loneliness and skill atrophy in this rapidly evolving landscape.
Key insights
AI tools enable 8x code output, shifting engineering focus to ambition, verification, and product sense.
Principles
- High agency in AI-native teams demands high accountability for outcomes.
- A growth mindset and continuous learning are vital for adapting to rapid AI advancements.
- Proactive quality measures and early detection are critical for managing increased code velocity.
Method
Implement AI routines for automated feedback analysis, code review, and PR generation. Adopt "just-in-time" monthly planning and embed "what good looks like" specs for AI-driven validation.
In practice
- Leverage AI agents to summarize feedback channels and draft initial bug fixes or feature PRs.
- Organize "pairwise programming lunches" to foster collaboration and share AI workflow techniques.
- Consistently use your team's product ("dogfooding") to gain firsthand user experience and identify issues.
Topics
- AI-driven Development
- Engineering Management
- Claude Code
- AI Routines
- Software Productivity
- Team Culture
- Product Sense
Best for: CTO, VP of Engineering/Data, AI Product Manager, AI Engineer, Director of AI/ML, Software Engineer
Related on AIssential
Editorial summary, takeaway, and curation by AIssential. Original article published by Lenny's Podcast: Product | Career | Growth.