CAPABILITIES.md

Not a skill list. A capability graph.

Capability Map

System Decomposition Under Pressure

Action pattern: Take a failing or overloaded system, identify the critical bottleneck, and redesign the architecture to eliminate it — while the system is still running in production.

Context: Production systems at scale, time-critical situations, distributed architectures.

Autonomy level: Full — can do independently, end to end.

Evidence: See DECISIONS.md — Payment System Migration (shadow mode approach).

AI amplification: AI coding assistants let me prototype architectural alternatives 3x faster, evaluating more options before committing.

Translating Between Technical and Business Languages

Action pattern: Take a complex technical constraint and reframe it as a business trade-off that non-technical stakeholders can evaluate and decide on.

Context: Cross-functional meetings, architecture reviews with leadership, investor communications.

Autonomy level: Full.

Evidence: Regularly present technical decisions to C-suite. The "shadow mode" framing for the payment migration is a direct example.

Building High-Trust Engineering Teams

Action pattern: Create an environment where engineers feel safe to challenge assumptions, admit uncertainty, and prioritize learning over looking competent.

Context: Teams of 5-15, growth-stage companies, technically complex domains.

Autonomy level: High — need occasional input from HR/people ops on policy.

Evidence: See DECISIONS.md — Hiring Junior over Senior. Team retention rate: 95% over 2 years.

Capability Combinations

My rare combination: Technical depth + stakeholder communication + team building

This matters because: Most people who can go deep on distributed systems can't present to a board. Most people who present well can't debug a race condition. I do both, which makes me effective as the bridge between engineering and business.

Tools & Methods

Category	Tools / Methods	Proficiency
AI & Automation	Claude, Cursor, GitHub Copilot	Daily driver
Development	TypeScript, Go, Python, PostgreSQL	Expert
Architecture	Event-driven, microservices, observability	Expert
Communication	Technical writing, architecture diagrams	Advanced

Learning Velocity

Recent example: Learned Rust in 3 weeks to prototype a performance-critical service. Shipped to production in 6 weeks.

My learning pattern: I learn by building. Give me a real problem and a deadline, and I'll learn whatever technology is needed to solve it.

Last updated: 2025-11-15

Example HCP Profile