Writing

On building AI infrastructure,
operating in capital-intensive industries,
and the capital structures that make both possible.

Long-form essays on the three disciplines I work across. Published infrequently. Each piece is something I am willing to defend in person.

01 — The Tracks

Three tracks. 8 essays. Published when each is something I'm willing to defend.

Track A — Energy & AI Infrastructure

On the software layer the renewable industry has been waiting for, and what the incumbents are about to miss.

Renewable O&M is the next great enterprise software market — and the incumbents are going to miss it.

The renewable industry has spent twenty years building assets and almost no time building the software to operate them. The existing players are dashboards. The next layer is agentic. Why this matters now, what changes for asset operators, and the three-year window before category consolidation closes.

Forthcoming

The four reasons solar operators don't trust AI — and what changes in 24 months.

Trust is the binding constraint on AI adoption in renewable operations, not capability. Asset operators have been burned by analytics vendors promising self-healing and delivering CSV exports. The unlock is edge-deployed agents that act locally and explain themselves. Walk through the trust-building sequence.

Forthcoming

Multi-manufacturer fleets are the unsexy reason renewables don't scale — and the AI opportunity hiding inside them.

Every utility-scale operator owns inverters from four to seven OEMs across a portfolio. Each speaks its own protocol, has its own portal, its own alarm taxonomy. The labor cost of simply understanding what is broken is enormous. The integration layer is where AI earns its keep — not in prediction, in translation.

Forthcoming

Track B — Operating in Private Equity

On the playbooks that get acquired companies from "deal closed" to "value created."

Why 90% of MSP roll-ups stall at 15% EBITDA — and the operational gates that get you to 25%.

MSP roll-ups under-deliver because acquirers buy revenue and inherit chaos. The 25% EBITDA target requires four specific gates — tooling consolidation, ticket taxonomy, automation layer, and incentive realignment — executed in a specific order. Most operators do them out of order or skip one. The playbook, with the order that matters.

Forthcoming

Technical debt is a P&L line item. Here's how to actually price it.

Every PE deck claims they'll "pay down technical debt." Almost nobody quantifies what it costs. How to measure technical debt in dollars — engineering hours diverted, defect carrying cost, delivery latency penalty, customer churn attribution — using AuthenticID and Ford as case studies.

Forthcoming

The fractional-CEO playbook for distressed AI/SaaS assets: a nine-month operating sequence.

Post-funding cleanup, post-acquisition integration, and post-founder transition are different problems with different playbooks — but most operators run the same generic plan on all three. The three sequences distinguished, with the inflection points that mark when one becomes another.

Forthcoming

Track C — Capital & Structure

On the corporate structures, tax tools, and capital instruments that determine what is possible to build.

ITC/PTC transferability changed renewable project finance. Most operators haven't updated their capital structure to match.

The Inflation Reduction Act's transferability provisions created a liquid market for energy tax credits. This fundamentally changes how renewable infrastructure platforms should be capitalized — what changes about JV economics, anchor customer terms, and exit timing. Walk through the new structure operators should be using by now and most aren't.

Forthcoming

Why every AI infrastructure company should be a multi-entity, not a single corp.

AI infrastructure companies in regulated industries have three distinct risk-and-reward profiles inside them — IP, operations, and project finance. Stuffing them into one C-corp creates tax inefficiency, exit friction, and investor confusion. The entity stack that actually works, and why most founders structure into ineligibility on day one.

Forthcoming

02 — Elsewhere

External work and appearances.

Interviews & Podcasts

Forthcoming as new appearances are confirmed.

In the press

Forthcoming as coverage publishes.

Speaking

Forthcoming as the speaking slate develops.

03 — Stay With It

New essays, occasionally.

I publish when a piece is finished, not on a schedule. If you'd like the next essay in your inbox when it goes live, leave your email below. No newsletter, no marketing — just the essay link when it publishes.

One email per new essay. Unsubscribe in one click.