Fractional CTO and advisor for seed-stage founders. Two decades bridging experimental R&D and production-grade systems at the seam of local AI, edge hardware, and agentic workflows.
I help seed-stage startups bridge the gap between experimental R&D and production-grade reality. My focus sits at the intersection of local AI models, edge hardware, and autonomous agentic workflows. Twenty plus years across networks, large-scale data, and high performance computing, with roots that go back to a USAF systems engineering role in Alaska.
I spent five years at Microsoft, most recently as Director of AI Strategy for Fortune 10 device partners, and before that as Director of Partner Technology Strategy for machine learning and AI across the Azure ecosystem. Across that tenure my work contributed to more than six billion dollars in partner solutions revenue. Earlier I spent two and a half years as a Big Data Solutions Architect at AWS, served as a Data Solutions Engineer at Databricks, and was CTO of Next Rev Technologies where I grew the P&L from 450K to 2.5M.
Today I run Qylogix as founder and CEO, advise a small portfolio of seed-stage AI and infrastructure companies, and serve as Director of Emerging Technologies at Hygia. I maintain a working research lab for distributed inference at home and occasionally end up in front of a classroom or a podcast microphone.
Architecture review, AI roadmap, technical hiring, partner introductions. The decisions a five-person team needs to get right before they have a fifty-person team. Engagements typically run six to twelve months.
Engagement details · qylogix.ai →Active advisor to VitalSpark AI (motivation and health platform) and Besample (academic research infrastructure operating across 40+ countries). Previously Managing Partner and angel investor at The Batchery, a Berkeley-based accelerator.
Early stage healthcare. Building the technical foundations and strategic narrative for a category that hasn't fully named itself yet. Working alongside the founding team.
Director of AI Strategy for Fortune 10 device partners. Director of Partner Technology Strategy across the Azure ecosystem. Generated 900M in solutions revenue. Azure OpenAI champion across the Microsoft partner ecosystem.
A field manual for big data processing and analytics using nothing but the tools already on your machine. Pipes, awk, jq, and a stubborn belief that text is still the universal interface. Co-authored with Christopher McCubbin and Raymond Page.
Speaker at the Yale School of Management's annual gathering of PE and VC practitioners. February 2025, NYC. On the operational and capital implications of AI infrastructure for portfolio companies.
Fireside conversation on the trajectory of AI hardware, devices, and the strategic implications for enterprise. Meet the Ghosts of Technologies Past, Present, and Future.
A CyberTangent podcast episode on what the modern CTO actually does. Less code review than people assume, more translation between worlds than anyone tells you about.
A wide ranging conversation on advances in AI hardware acceleration, OEM device strategy, and the Batchery accelerator's portfolio approach to applied AI.
On the discipline of treating privacy as architecture rather than afterthought, and why scope discipline is the cybersecurity skill nobody teaches.
Twenty years designing reference architectures across AWS, Azure, and bare metal for partners ranging from Fortune 10 down to seed-stage. The job is rarely recommending a stack. It's knowing what NOT to build, when to buy versus build, and which architectural debt is worth taking on at your specific stage.
As an angel investor and former Managing Partner at The Batchery, I've read enough subscription agreements to know what kills a deal. I can read your investor docs as fluently as your architecture diagrams. Useful when the term sheet shows up before the technical due diligence does.
Twenty years of working relationships across Microsoft, AWS, Databricks, NVIDIA, Cisco, Google Cloud, and Oracle. When a portfolio company needs a partner intro, a hyperscaler co-sell, or an OEM integration, those calls actually get returned.
I write code, ship hardware, and run production systems. The day I stop is the day I should stop advising founders on technical decisions. The distance between PowerPoint and a running cluster is where most fractional engagements quietly fail.