“Face the sunshine, and forge a path in the uncharted wilderness.”

This manual is designed to reduce communication friction and help us collaborate effectively. Whether as a founder, technical partner, or team builder, my underlying engineer’s soul remains unchanged. I went from firmware engineer to leading a 250+ person enterprise R&D organization, and today I build products hands-on as a one-person company working with AI agents. Whether you’re on my team, a partner, or a future colleague or manager, this manual applies. This is a living document reflecting my current operational state.

🌟 Core Philosophy & Vision

  • Embracing Ambiguity & Action Speaks: When exploring the frontiers of emerging tech (like AI, Robotics, and Cyber-Physical Systems), the map is rarely drawn out. I deeply value resilience and tangible execution when facing the unknown. Tell me what we can do, rather than just pointing out what’s broken.
  • Optimism: I’m optimistic by nature. When setbacks or bad news hit, my first question is “what can we do next?” rather than dwelling on blame or complaint. That’s also why I want bad news straight away: the sooner I know, the better our chances of solving it together.
  • Fairness Cares: I deeply value fairness. I will intervene if team members are treated unfairly or if actions negatively impact the collective good.
  • Empowerment & Psychological Safety: I detest micromanagement. I provide context and expect you to take the steering wheel. We are navigating the unknown together; honest mistakes are an acceptable part of exploring the frontier. Watching you take ownership and grow is my greatest satisfaction.
  • Coaching: As an ICF-certified coach (ACC), I ask questions rather than hand out answers, so that each person becomes the owner of their own domain. The biggest thing the credential gave me, honestly, was learning to coach myself first.

⚙️ Technical Philosophy & Dev Culture

  • Architectural Thinking (DDD): I have a strong bias toward Domain-Driven Design (DDD). I appreciate systems that are structured, loosely coupled, and accurately reflect real-world business realities.
  • Exploring Technical Frontiers: Shipping the feature is only the baseline; pushing boundaries is the goal. I encourage deep curiosity and practical exploration of edge tech (e.g., Rust, edge computing, WebTransport, MCP, and AI agents).
  • AI-Driven Workflows: I actively embrace AI and expect teams to use AI tools to raise their output substantially. But once starting gets cheap, the bottleneck moves to knowing whether the work is right: check AI conclusions against their sources, turn rules written in documents into tests, and whatever can actually be run, run it. Human brainpower goes to verification, trade-offs, and architectural decisions.

💬 Communication & Working Preferences

  • Async & Efficiency First: I prefer crisp documentation over lengthy, inefficient meetings: important decisions go into numbered decision records, and every stretch of work ends with a retrospective. For urgent matters, though, a direct call or face-to-face conversation always beats email or back-and-forth messages.
  • Be Straight & Open Minded: I prefer difficult, honest conversations over unspoken issues. I view all feedback as a learning opportunity; please be direct and offer specific suggestions.
  • Deep Processing & Quiet Concentration: As an introvert, I work best in quiet environments for solving complex problems. If I am silent in a meeting, it rarely means I disagree—it means my brain is compiling in the background to find the optimal solution. When we need to talk something through, just set up a time.

🤖 Working with My AI Agent Team

These days I collaborate not only with people, but with an army of AI Agents. The principle is the same as with people: I own judgment; AI exists to expand my imagination, reduce my blind spots, and accelerate experiments. AI Agents may propose, challenge, simulate, implement, critique, and test — but they don’t replace my judgment. My job isn’t to accept AI output; it’s to shape that output into product value.

  • Manage the tension with four modes: I use Explore / Create / Build / Refine to balance elegant architecture against fast iteration. The system fails if it only builds, and fails if it only explores. The goal is creative execution.
  • Watch for fake progress: more documents but less clarity, more code but less value, more agents but less judgment — all signals to stop. AI should increase creative leverage, not breed a “bureaucracy of one.”
  • A solo founder needs psychological safety too: as with a team, I treat delays, pivots, and bugs as learning rather than failure — challenged, not threatened.
  • Turn lessons into rules: every stretch of work ends with a retrospective, and the pitfalls go into the agents’ rules file: state only confirmed facts; check a subagent’s key verdicts against the source before relaying them; and in a direction review, ask not only “was it built?” but also “is this the method I want?”

The full operating philosophy is here: Solo Founder × AI Agent Army: My Operating Directive.