People to follow for compound engineering, context engineering, and AI agent development.
Compound Engineering & AI Agents
Vinci Rufus
Website: vincirufus.com
Focus Areas:
- Agentic AI systems and multi-agent architectures
- Compound engineering methodologies
- AI agent design patterns and reliability
- Context engineering and prompt optimization
- Production-ready AI deployment
Key Articles:
- The Ralph Loop and Compound Engineering – Fresh context iteration patterns
- Compound Engineering – The productivity equation:
Productivity = (Code Velocity) × (Feedback Quality) × (Iteration Frequency) - Agent Loop Architecture – Seven components of agent loops
- The Reliability Chasm in AI Agents – Why 95% of PoCs fail in production
Philosophy: “Believes in Agentic AI” – focuses on building swarms of intelligent agents
Notable themes: Business-to-Agent (B2A) SaaS, the evolution of software economics through AI, the distinction between demo agents and production systems, Model Context Protocol standardization.
Presence: Active on GitHub and LinkedIn with 40+ technical articles
Why Follow: Practical implementation focus over theory. Addresses real challenges like reliability, evaluation systems, and the transition from conversational to truly agentic interfaces.
Dex Horthy (HumanLayer)
Website: humanlayer.dev
Focus Areas:
- Human-in-the-loop agent design
- Production agent architecture
- Agent approval workflows
Key Work:
- 12-Factor Agents – Principles for production-ready LLM software
Why Follow: Deep practical knowledge of agent deployment challenges and human-AI collaboration patterns.
Geoffrey Huntley
Focus Areas:
- RALPH Loop methodology originator
- Agent orchestration patterns
- Developer tooling
Why Follow: Original creator of the RALPH Loop concept that enables fresh-context agent iteration.
Steve Yegge
Focus Areas:
- “Gas Town” multi-agent future concept
- Software engineering evolution
- Developer experience
Why Follow: Thought leader on how multi-agent systems will reshape software development.
Will Larson (Lethain)
Website: lethain.com
Focus Areas:
- Engineering leadership and management
- Infrastructure at scale
- Systems thinking
Key Work:
- Learning from Every’s Compound Engineering – Analysis of the Plan → Work → Review → Compound cycle
- Staff Engineer, An Elegant Puzzle (books)
Why Follow: Brings engineering leadership perspective to compound engineering. His analysis emphasizes that “the quality of your codebase, tests, and continuous integration harness” matters more than the agent itself.
Anthropic Engineering Team
Website: anthropic.com/engineering
Focus Areas:
- AI safety and alignment
- Claude capabilities and best practices
- Production AI deployment patterns
Key Work:
- Effective Harnesses for Long-Running Agents – Initializer + Coding Agent architecture, JSON feature lists, session initialization sequences
Why Follow: First-party guidance on building robust agent systems with Claude.
How to Use This List
- Subscribe to RSS/newsletters – Stay current on new articles
- Read foundational pieces – Articles linked above are canonical
- Cross-reference with knowledge base – See how concepts connect to your own documentation
- Track emerging patterns – New methodologies often appear here first
Adding to This List
When adding new thought leaders:
### [Name]
**Website:** [url]
**Focus Areas:**
- Area 1
- Area 2
**Key Work:**
- [Article Title](url) - Brief description
**Why Follow:** One sentence on their unique value.
