Software Engineer | AI-Assisted Development | Systems Thinker
I'm a software engineer based in Auckland, specializing in using AI coding agents as part of real-world engineering workflows. With 30+ years of systems-level experience and 8+ years of intensive Python development, I focus on architectural reasoning, agent orchestration, and building production-ready systems.
I treat coding agents as force multipliers, not replacements for engineering judgment:
- Spec-first development - Write clear, testable specifications before delegating to agents
- Multi-agent orchestration - Assign distinct roles (exploration, refactor, validation, documentation)
- Supervised iteration - Review and refine agent output to maintain system integrity
- Human-in-the-loop - Know when to rely on agents vs when to intervene manually
I use Cursor daily for long-context prompting, code generation, and supervised refactoring of production codebases.
Languages & Frameworks:
- Python (FastAPI, Flask, SQLAlchemy, async programming, pandas)
- SQL (PostgreSQL - schema design, optimization, complex queries)
- Bash scripting & Linux systems administration
AI & Agent Tooling:
- Cursor (extensive daily use)
- OpenAI, Anthropic, Ollama (local models)
- Function calling, structured prompting, spec-driven generation
Infrastructure:
- Docker & containerized deployments
- Linux (daily driver), SSH-based workflows
- Nginx, reverse proxies
- Git for collaborative development
EventHub - Production full-stack platform built with AI-assisted development
This project demonstrates my agent-centric engineering approach:
- PostgreSQL schema with 15+ related tables designed with agent assistance
- Async FastAPI backend with comprehensive API surface
- AI-powered event creation using LLM function calling (GPT-4, Claude, local models)
- OAuth2 authentication, Stripe payments, webhook handling
- HTMX frontend with progressive enhancement
- Dockerized deployment with zero-downtime updates
Agent Workflow Highlights:
- Used agents to explore and extend the growing codebase
- Wrote detailed specs before delegating implementation
- Employed agents for schema reasoning and refactor safety checks
- Iteratively reviewed agent output to maintain architectural integrity
Systems-Level Foundation: 30+ years spanning professional programming (started at 16), Linux/Unix administration, database architecture, and hardware diagnostics. This long arc provides strong intuition about failure modes, edge cases, and system behavior.
Real-World Engineering: Built production tools and analytical systems in insurance operations, bridging business requirements with technical implementation. Experience reasoning about messy, real-world systems with ambiguous requirements and human factors.
Agent-Supervised Development: Extensive experience using AI coding agents to debug, refactor, and extend non-greenfield codebases while maintaining human control over architecture and correctness.
- Systems thinking - Architecture, data models, failure modes, trade-offs
- Agent orchestration - Decomposing complex problems, selecting appropriate agent roles
- Debugging unfamiliar code - Tracing logic/state issues, incremental fixes
- Asynchronous engineering - Clear written communication, reproducible reasoning
- Long-context reasoning - Effective in ambiguous requirements and complex codebases
- Email: laptoprecording@gmail.com
- Location: Auckland, New Zealand (Remote-friendly, async timezone)
- Availability: 10-20 hours/week initially, comfortable scaling up
Treating agent output as proposed diffs, not authoritative truth. Check out my repositories to see AI-assisted development in practice.