Operators on the stage.
Not consultants who left the field.

Most companies bolted AI onto their workflow and watched the bolt rust off in a quarter. We treat AI as infrastructure — installed once, owned by your team, governed by standards that survive when the model APIs change shape. This talk walks through the anatomy of a real operator AI stack: what gets installed, who owns each layer, and how to tell a workflow from infrastructure.

Audience walks away with: the difference between a workflow and a system · a six-layer reference architecture · a self-audit framework for their current stack · the four patterns that make skills survive a model change.

Best for: AI conferences · operator summits · CTO/CEO audiences

The consulting industry is being rebuilt from the bottom up. Hourly billing is a tax on the buyer. Deliverables are the new product. We turned our own operating system into a productized consultancy — fixed scope, fixed price, code as deliverable, framework library as moat. This talk covers what changed, what's coming, and how services firms either rebuild or get squeezed between AI tools and full-service agencies.

Audience walks away with: a framework for productizing any service · a pricing model that survives commoditization · how to build a moat that isn't talent · the three ways AI breaks the consulting business model.

Best for: agency / services conferences · founder communities · consulting firms

Two operators running five businesses — that math only works if the leverage compounds. We'll walk through the skill library that runs underneath everything we touch: nightly media optimization, hourly comment marketing, daily data validation, weekly funnel reports, monthly close support. How each one was built, what it replaced, what it costs, and why the seventh skill is exponentially easier than the first.

Audience walks away with: the COLLECT → FILTER → BATCH → EXECUTE pattern that makes skills cheap to run · the skill-authoring standard that makes skill N+1 trivial · a real cost breakdown of running 60+ skills · the three roles every skill needs to outlast its author.

Best for: small-team founder conferences · ops/automation events · ecom communities

Most brand books are PDFs that no one reads after the rebrand party. We treat brand as a runtime system — voice rules that lint your copy, design tokens that compile your CSS, audience cohorts encoded as enforceable language constraints. The result: every piece of customer-facing work passes through the brand standard automatically, even when we're not in the room. This talk shows how that works and why it's the only way brand survives an AI-content explosion.

Audience walks away with: what to operationalize in a brand book vs. what to leave aspirational · the structure of voice rules that AI can actually enforce · a copy-rules taxonomy with examples · how brand becomes a moat in the era of infinite content.

Best for: brand / creative conferences · marketing leadership · agency events

Sized to a specific buyer: too big for the consumer SaaS tools, too small for the McKinsey transformation deck. We've spent the last three years operating in this exact band. This talk covers what AI actually moves at this scale (margin first, growth second), where the leverage points hide (data integrity, creative production, retention), and what to ignore (most of what's on LinkedIn). Includes a 90-day adoption sequence we've run inside our own businesses.

Audience walks away with: a capability map for consumer brands at scale · the order to install AI in (and what to avoid touching first) · a reference list of the 12 highest-leverage skills · the 3 traps that kill 80% of mid-market AI initiatives.

Best for: ecom / DTC conferences · Shopify summits · retail / consumer events

A counter-positioned talk. Every AI talk is a highlight reel; ours isn't. We'll walk through the implementations that quietly stopped getting used, the budget we burned on autonomous agents that hallucinated their way through customer support, the tool we shipped to four brands that none of them touched after month two, and the one we keep installing because it's the rare thing that compounds. Most useful talk we give. Audiences ask the most questions about this one.

Audience walks away with: the four-failure taxonomy we use to triage AI projects · how to spot a "haunted demo" before signing · a kill-criteria template every internal AI initiative should have · the operator's case for human-in-the-loop, even when fully autonomous works.

Best for: anywhere with a sophisticated audience · executive forums · CTO / CEO peer groups