Iggy M.

Check my work in fifteen minutes.

I'm Iggy M., an AI agent architect in Dallas, Texas. I build the infrastructure that makes AI agents reliable in production. This page is for hiring managers: where to look, the full career record, and how to reach me.

Fifteen minutes, four stops

  1. 3 min

    Run a crash test

    Open CRASH/TEST, pick any of the ten incidents and run it bare. Then fit one control and run it again. It shows how I think about agent failure: what broke, and which control would have stopped it.

    Open CRASH/TEST

  2. 5 min

    Point it at your own setup

    Paste your team's agent configuration into the checker. It lists which of the ten incidents your setup would stop, and which rest on a written rule alone.

    Open the checker

  3. 5 min

    Read some code

    Open repositories covering guardrails for agent state transitions, a deterministic supervisor, token management and interpretability tooling.

    factory-gate circuit-x deterministic-agent-supervisor sliding-window-trim-engine

  4. 2 min

    Ask the hard questions

    The Ship Mind answers from my career record and nothing else. Ask it what you would ask in a phone screen.

    Interrogate the Ship Mind

Career record

  1. 2024 to now

    Independent / Nubla AI & open work

    AI Agent Architect & Software Factory Creator

    Currently building the systems that turn frontier AI into reliable, autonomous engineering leverage. This includes multi-agent orchestration pipelines, software factory infrastructure, deterministic supervisors that prevent LLM-induced errors, context engineering toolkits, sliding-window token management, and interpretability tools (such as CircuitX) that expose internal feature circuits for alignment. The goal is simple: make one engineer as productive as a team by engineering the infrastructure that works without constant supervision.

    • Created factory-gate: autonomous structural policy guardrails that verify safe agent state transitions before code reaches target branches
    • Built CircuitX for mechanistic reverse-engineering and interpretability-driven alignment of deep neural weights and code LLMs
    • Engineered deterministic-agent-supervisor and sliding-window-trim-engine for reliable, efficient agent workflows
    • Develops software factory systems and AI harness layers that connect models, tools, memory, evaluation, and human oversight into production-grade pipelines
  2. 2022 to 2023

    Apple

    Cloud Architect and Engineer

    Worked on cloud architecture and engineering inside one of the world's most demanding consumer platforms. The environment required absolute reliability, privacy-preserving design, and systems that serve hundreds of millions of devices without drama. Focused on infrastructure that stays invisible when it works and fails gracefully when it doesn't.

    • Designed and delivered cloud services and infrastructure supporting large-scale consumer products where uptime and polish are non-negotiable
    • Applied systems thinking honed at carriers to Apple-scale problems: latency budgets, failure domains, and operational simplicity
    • Contributed to architectures that balance security, performance, and developer experience at extreme scale
  3. 2017 to 2021

    Sam's Club

    Principal Software Engineer — Cloud and Machine Learning

    Principal engineer focused on cloud platforms and machine learning systems inside a major retail operation. The role sat at the intersection of large-scale data, production ML, and cloud infrastructure — the practical foundation for later work on AI agents and software factories. Emphasis on systems that deliver measurable business value while remaining maintainable and observable.

    • Led cloud and machine learning engineering initiatives that brought production ML into core retail workflows
    • Designed scalable cloud architectures supporting data-intensive and ML-powered features
    • Bridged research-grade models with operational reality: monitoring, versioning, reliability, and cost control
  4. 2007 to 2009

    AT&T / Verizon

    Mobile and Web Applications Architect Lead (AT&T); SOA Designer and Architect (Verizon)

    Led mobile and web application architecture at AT&T and contributed SOA design work at Verizon. This was carrier-grade territory: networks where uptime is measured in nines, change management is deliberate, and systems must survive real-world conditions. The work built a deep intuition for distributed systems, service boundaries, and operational rigor that later proved essential for cloud and AI systems.

    • Architected and led mobile and web application platforms at AT&T serving carrier-scale traffic
    • Designed SOA architectures at Verizon, establishing clean service contracts and integration patterns
    • Developed habits of reliability engineering, careful change management, and thinking in failure domains that remain core to later work

Newest first. These are the periods on file; ask about anything in between.

How I design systems

  • Edge-first and latency-aware design: run logic and inference where the users and data already are
  • Resilience by default: graceful degradation, deterministic guardrails, and failure that is boring rather than catastrophic
  • Observability, security, cost, and interpretability treated as non-negotiable requirements at every layer
  • Software factory thinking: engineer the pipelines and supervisors so that AI can plan, execute, verify, and iterate without creating traffic jams

Working with me

  • Has led architecture and engineering efforts across mobile, web, cloud, and now AI systems at companies ranging from carriers to Apple to retail
  • Mentors through concrete artifacts: code, diagrams, and working systems rather than abstract advice
  • Works effectively with stakeholders by grounding technical decisions in latency, reliability, cost, and measurable outcomes

What I work with

Domains
Multi-agent orchestration & software factories. AI harness & context engineering. Edge & distributed systems. Production ML & reliability. Interpretability & alignment tooling. Networking & security architecture. Observability & cost engineering.
Platforms
Cloudflare (Workers, Pages, R2, D1, KV, Zero Trust), AWS, GCP, Azure, Kubernetes, Docker, GitHub Actions.
Languages
Python, TypeScript, Go, Swift, Java, Rust, JavaScript, C#.

Résumé

This page has a print layout. Print it, or save it as a PDF, and you get the career record as a résumé. For applicant-tracking software and agents, the same record is at /hire/resume.json in JSON Resume format.