Selling outcomes in the age of AI infrastructure

My career has followed the infrastructure that powers the economy. I spent about ten years at AWS, then ran revenue at Electric Era, a battery-backed EV fast-charging company, and now lead the commercial business at Patmos, which builds AI campuses and offers colocation, cloud and hosting services.

The common thread is that customers do not buy infrastructure. They buy what it lets them do.

Sell outcomes, not specifications

Infrastructure companies love specifications. Customers, especially executives, care about outcomes: speed to market, reliability, cost and risk. A solutions catalog that turns technical components into customer-ready offers is one of the highest-leverage things an infrastructure company can build.

Reliability is a product

At Electric Era, I sold reliability as the core differentiator: a financially backed 98% uptime commitment, with the network achieving 99.6% uptime. When reliability is backed by a commitment, it stops being a marketing claim and becomes something a customer can underwrite.

Speed is a product too

In infrastructure, time to live service is often worth more to the customer than a lower price.

Move from licenses to consumption

At AWS, I helped a strategic healthcare partner win competitive bids against the incumbent, and helped its AI/ML unit cut model-training TCO by more than 50%. The partner's revenue grew 6x in two years. Consumption models align what the customer pays with the value they get, which is exactly what AI workloads demand.

Why this matters now

The AI buildout is creating a wave of infrastructure companies, from compute to power to software platforms. The ones that win commercially will be the ones that package outcomes, back their claims, and build commercial controls strong enough to support long-lived, capital-intensive commitments.

Commercial controls for long-lived commitments

Infrastructure deals are different from software deals. Commitments run for years, the capital at stake is large, and delivery depends on physical work, supply chains and power. That makes commercial controls part of the product. Every new offer needs a cost basis before it's priced. Delivery needs to sign off on scope, cost and lead time before a contract is signed. Terms need to be checked against ranges agreed with Finance, and the largest commitments need named approvers.

Customers notice this discipline. A provider that knows its costs and honors its delivery dates is easier to trust with a long commitment.

Integrate with how the customer already works

At Electric Era, we positioned charging as retail infrastructure, not just electricity. That meant integrating stations with retailer loyalty, point-of-sale, CRM and retail-media programs, and helping launch RetailEdge, an AI-enabled retail charging kiosk, at the NACS Show. Infrastructure becomes much easier to buy when it plugs into systems and goals the customer already has.

Explain the "why" in public

Part of selling infrastructure today is making the case in public. I've spent much of 2026 on national television and radio talking about the AI race, American competitiveness and why AI and data center infrastructure should be built in American communities. Buyers, regulators and communities all shape whether infrastructure gets built, and commercial leaders need to be able to speak to all of them.

Each audience needs a different version of the same story. Buyers want outcomes, investors want discipline, and communities want to know what the project means for them. A commercial leader who can tell all three versions honestly makes the whole company easier to say yes to.

That is a big part of why I enjoy this work.

That is the work I do.