How this PM Used Claude Code to Support 20 People

The future of product management in an AI-native era centers on PMs supporting broader roles with blended functions, necessitating well-organized, high-context knowledge repositories. Hannah Stulberg, a PM at DoorDash and former Google APM, introduces the "Team OS," a team knowledge base built in Claude Code, emphasizing context management through lean Claude MD files and structured information. The Team OS enables efficient data analysis, bug investigations, and cross-functional collaboration by using shared agents, commands, and skills. Stulberg stresses the importance of planning with Claude, using plan mode to refine strategies and create high-quality documents. She advocates for a beginner's mindset, encouraging continuous learning and adaptation to AI tools, and highlights that Claude Code is not just for coding but for all knowledge work.

Outlines

Part 1: The Evolving PM Role

Part 2: Team OS and Context Management

Part 3: Scaling Collaboration and Data

Part 4: Advanced Planning and Claude Optimization

Part 5: Mindset, Tools, and Future Growth

Sign in to continue reading, translating and more.

Continue

Aakash Gupta

Part 1: The Evolving PM Role

The Evolving Role of Product Managers: Supporting Broader Teams and Blended Functions

Part 2: Team OS and Context Management

Team OS: A Centralized Knowledge Base for Enhanced Team Collaboration and Context Sharing

Context Management: Optimizing LLM Performance Through Lean Claude MD Files and Thinking Room

Navigating the Team OS: Efficiently Querying Customer Data with Doc Indexes and Contextual Awareness

Structuring Information for LLMs: Shared Agents, Commands, and Skills for Consistent Data Synthesis

Part 3: Scaling Collaboration and Data

Scaling Data Analysis: Leveraging the Team OS for Metrics, Queries, and Verified Approaches

Collaborative Knowledge Repository: Empowering Engineers and Streamlining Feature Launches

Integrating Daily Product Work: Collaborative Context Sharing and Pull Requests for Team OS

Overcoming GitHub Fears: Natural Language Commands and Streamlined Pull Request Processes

Part 4: Advanced Planning and Claude Optimization

Enhancing Document Quality: Planning Effectively with Claude and Understanding Plan Mode

Optimizing Claude's Performance: Minimizing Token Usage and Parallelizing Research

Advanced Planning Techniques: Verification, Strategy Doc Structure, and Thinking Partner

Refining Plans: Inviting Claude as a Thinking Partner and Leveraging Writing Guides

Part 5: Mindset, Tools, and Future Growth

Embracing a Beginner's Mindset: Learning and Improving with Claude's Guidance

Overcoming Early Frustrations: The Importance of Persistence and Continuous Learning

Choosing the Right Tool: Coding Agents for Advanced PM Work and Automating Time

Underhyped Aspects: Curiosity, Expertise, and the Misleading Name of Claude Code

Overcoming Terminal Fears: Embracing a Beginner's Mindset and Connecting MCPs

How this PM Used Claude Code to Support 20 People

Aakash Gupta

Part 1: The Evolving PM Role

01:45The Evolving Role of Product Managers: Supporting Broader Teams and Blended Functions

The Evolving Role of Product Managers: Supporting Broader Teams and Blended Functions

Part 2: Team OS and Context Management

03:11Team OS: A Centralized Knowledge Base for Enhanced Team Collaboration and Context Sharing

Team OS: A Centralized Knowledge Base for Enhanced Team Collaboration and Context Sharing

05:50Context Management: Optimizing LLM Performance Through Lean Claude MD Files and Thinking Room

Context Management: Optimizing LLM Performance Through Lean Claude MD Files and Thinking Room

08:11Navigating the Team OS: Efficiently Querying Customer Data with Doc Indexes and Contextual Awareness

Navigating the Team OS: Efficiently Querying Customer Data with Doc Indexes and Contextual Awareness

13:16Structuring Information for LLMs: Shared Agents, Commands, and Skills for Consistent Data Synthesis

Structuring Information for LLMs: Shared Agents, Commands, and Skills for Consistent Data Synthesis

Part 3: Scaling Collaboration and Data

15:21Scaling Data Analysis: Leveraging the Team OS for Metrics, Queries, and Verified Approaches

Scaling Data Analysis: Leveraging the Team OS for Metrics, Queries, and Verified Approaches

21:02Collaborative Knowledge Repository: Empowering Engineers and Streamlining Feature Launches

Collaborative Knowledge Repository: Empowering Engineers and Streamlining Feature Launches

25:28Integrating Daily Product Work: Collaborative Context Sharing and Pull Requests for Team OS

Integrating Daily Product Work: Collaborative Context Sharing and Pull Requests for Team OS

29:07Overcoming GitHub Fears: Natural Language Commands and Streamlined Pull Request Processes

Overcoming GitHub Fears: Natural Language Commands and Streamlined Pull Request Processes

Part 4: Advanced Planning and Claude Optimization

33:30Enhancing Document Quality: Planning Effectively with Claude and Understanding Plan Mode

Enhancing Document Quality: Planning Effectively with Claude and Understanding Plan Mode

38:41Optimizing Claude's Performance: Minimizing Token Usage and Parallelizing Research

Optimizing Claude's Performance: Minimizing Token Usage and Parallelizing Research

45:50Advanced Planning Techniques: Verification, Strategy Doc Structure, and Thinking Partner

Advanced Planning Techniques: Verification, Strategy Doc Structure, and Thinking Partner

53:05Refining Plans: Inviting Claude as a Thinking Partner and Leveraging Writing Guides

Refining Plans: Inviting Claude as a Thinking Partner and Leveraging Writing Guides

Part 5: Mindset, Tools, and Future Growth

59:22Embracing a Beginner's Mindset: Learning and Improving with Claude's Guidance

Embracing a Beginner's Mindset: Learning and Improving with Claude's Guidance

1:03:03Overcoming Early Frustrations: The Importance of Persistence and Continuous Learning

Overcoming Early Frustrations: The Importance of Persistence and Continuous Learning

1:04:08Choosing the Right Tool: Coding Agents for Advanced PM Work and Automating Time

Choosing the Right Tool: Coding Agents for Advanced PM Work and Automating Time

1:05:35Underhyped Aspects: Curiosity, Expertise, and the Misleading Name of Claude Code

Underhyped Aspects: Curiosity, Expertise, and the Misleading Name of Claude Code

1:07:24Overcoming Terminal Fears: Embracing a Beginner's Mindset and Connecting MCPs

Overcoming Terminal Fears: Embracing a Beginner's Mindset and Connecting MCPs