How the engineer behind Claude Cowork actually uses Claude | Felix Rieseberg (Anthropic)

2026-05-28 · Source: Lenny's Newsletter · Field: Technology & Digital — Artificial Intelligence & Machine Learning, Software Development & Engineering, Internet of Things (IoT) & Connected Devices · Depth: Intermediate, extended

Summary

Felix Rieseberg, Engineering Lead for Claude Cowork, Claude Code, and desktop applications at Anthropic, shares his advanced workflows and insights into utilizing Claude for both personal and professional tasks. He demonstrates building a \$20 hardware "Claude buddy" that provides physical notifications for AI actions, highlighting the potential for integrating AI beyond traditional interfaces. Rieseberg details using Claude Cowork to process personal documents like floor plans and emails for inventory, creating interactive 3D home planners and automated promise trackers. He also discusses the strategic choice between Claude Opus and Sonnet 4.6 models, emphasizing Sonnet's capability for well-scoped problems. The discussion extends to live artifacts for dynamic personal dashboards, the importance of providing context via connectors (e.g., Spotify, Gmail), and designing for asynchronous AI interactions by fostering user trust rather than constant supervision.

Key takeaway

For AI Product Managers designing new tools, prioritize user trust and asynchronous experiences over immediate, synchronous responses. Focus on enabling AI to handle tedious background tasks, freeing users for creative work. Your goal should be to empower users to define problems at a higher abstraction level, allowing the AI to autonomously determine the solution's implementation, even for novel hardware integrations or dynamic data dashboards.

Key insights

AI's true power lies in automating tedious tasks and enabling creative problem-solving, especially when users provide rich context and trust its autonomous execution.

Principles

• Almost any problem can be addressed by AI tools.
• AI should automate tedious tasks, not just cursor movements.
• Trusting AI's judgment enhances its problem-solving.

Method

To discover AI use cases, continuously abstract tasks, asking "How can Claude do this?" and "How can I never do this again?" Employ a "reverse interview" by describing your challenges to Claude and soliciting its solutions.

In practice

• Inventory personal assets using email as a data source.
• Automate promise tracking from messages.
• Create dynamic personal dashboards with live data.

Topics

• Claude Cowork
• Live Artifacts
• AI Hardware Integration
• Personal Productivity
• Human-AI Interaction
• LLM Model Selection

Best for: AI Engineer, AI Product Manager, Software Engineer

Related on AIssential

• 😸 Practical Workflows for working with Claude Cowork — Claude Cowork transforms AI from a chatbot into an active desktop agent capable of complex workflow automation.
• How to use Claude Cowork as a Competitive Intelligence System — Claude Cowork enables delegation of work and artifact creation through a desktop app, integrating with common work tools.
• Why Anthropic Thinks AI Should Have Its Own Computer — Felix Rieseberg of Claude Cowork & Claude Code Desktop — Claude Cowork extends Claude Code's capabilities to non-technical users through a user-friendly, VM-sandboxed environment.
• AI and Claude Design: Will it Replace Designers by 2026? — Claude Design applies Anthropic's AI, like Claude Opus 4.7, to design workflows, sparking debate on AI's role in design.
• Claude Dispatch and the Power of Interfaces — Suboptimal AI interfaces, especially chatbots, create cognitive load that limits user productivity and access to advanced AI capabilities.
• How to Maximize Claude Cowork — Claude Cowork offers a simplified, visual interface for Claude Code's capabilities, benefiting both technical and non-technical users.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by Lenny's Newsletter.