I work where technology, talent, business models, systems, and dealmaking have to become one thing that can actually run.
I work with founders, CEOs, investors and operating leaders when an important build has become bigger than a technology project.
Architecture is the mechanism. The outcome is eliminating the gap between the recommendation and the working system.
The material is the whole company, not the stack. Where the money moves, who decides, what gets owned, which parts were never built to fit.
Most business problems are structural problems disguised as feature problems. Rarely the code — usually a decision above it that nobody made. The feature ships. The structure underneath it stays exactly where it was.
Architecture tells you what to build. Assembly is how it actually gets built — talent, technology, business models, systems, and dealmaking. A term sheet is architecture. So is an operating model, and who ends up owning the thing. Every other architect draws. I assemble.
I find the break between the parts, architect the connection, and stay long enough to prove it in operation.
With AI, everyone is a Growth Engine. We’re all architects, producers, agents, builders, and owners of our IP.
These roles aren’t new. What’s new is that AI made all five available to one person at once — and almost nobody has the structure to run themselves that way yet. I know, because I spent ten years building the structure before I could name it.
You’re already running all five. The question is whether anything underneath you is built for it.
I find the decision above the feature, then build the structure every layer beneath it can inherit.
Innovation Labs at MCI. The first IBM SP2 out the door, and the largest NeXT purchase anyone had made. Then Discovery, Nokia, Charter, Nextel.
A good architecture still has to survive people, timing, tradeoffs, and the room where the work gets made.
The Blacklist writers’ room, 2015 to 2017. I went in to build software and spent two years watching a hundred-person production get assembled every week. The thesis started there.
Representation is architecture too: aligning the person, the opportunity, the terms, and the belief underneath them.
Apex Global. I’m the architect and the agent — which means I don’t just represent the people, companies, and ventures on the roster. I believe in them enough to build alongside them.
I stay where the interfaces meet — technology, talent, operating model, and the practical work of making them cohere.
Mediachase, 1997 to 2012. The dot-com crash killed the first company. I rebuilt it as a software company — ECF ran The Knot, Hershey, Teleflora, USPS — and sold that one. Optimizely is only now retiring the code, fourteen years after the exit.
The instance is temporary. The architecture compounds when you know what to keep, what to own, and what can run again.
The exit to EPiServer in 2012 carried 150 clients across. What I kept was better than what I sold: the architecture, not the instances.
Scroll through the five · or jump to one
Being early is easy to claim and hard to check. Seven of my eight calls are twenty years old — streaming in 1999, online marketplaces in 2000, commercial open source after that. All true, and all of it old. The fair question is what I’ve read correctly lately.
IP Companions™ is the answer. It doesn’t bet on a model — it defines the lifecycle slots where AI participates, and the provider plugs in. Mastra, Claude, OpenAI, whatever comes next. I architected the slots the models plug into. The model layer commoditized. The architecture did not.
That’s the eighth shift, and it’s the one I’m in.
I build, own, and stay long enough to leave the infrastructure behind. Each of these is a place where the technology problem and the operating problem turned out to be the same problem.
Where the IP lives. Creator Cloud™, IP-Lock™, and IP Companions™ were all built here.
The Open Talent Network™, operationalized as an agency. I built it, and I represent on it.
The Innovation Studio Framework™, deployed as a live studio. New York, Los Angeles, Stockholm. Three practices.
Enterprise delivery at scale. Senior advisory, and the parent of Ignite Studio.
I take one kind of engagement: a company carrying a build, a launch, or a transformation it can’t get over the line — where the problem reads as technology and turns out to be structure.
In 2001 the dot-com crash killed the business I’d spent four years building. I didn’t fold Mediachase. I rebuilt it as a software company, shipped the eCommerce Framework, and sold that one in 2012.
A company is an architecture, and architectures can be re-shaped without being abandoned. Most founders confuse the business they’re in with the company they built. Knowing which layer actually died is the whole skill — and it’s why I’m useful when a transformation goes sideways. Which it will.
Not a value statement. The deal structure.
The one that survived the strategy, the reorg, and the last five features. That’s where the connection layer went missing. Architecture is the only thing that changes the shape of the work from there.
A bad reason to call: a feature problem. I’m the wrong fit and it goes badly for both of us.
You can change the shape of the company, not merely approve another workstream.
You care about durable infrastructure, operating advantage, and whether the work compounds.
Your mandate outruns the systems you inherited, and you have room to redesign the way the work happens.
You are not buying a deck. You are looking for someone willing to help reshape and build the thing with you.
Send me the problem that keeps coming back. One paragraph is enough — the one that survived the strategy, the reorg, and the last five features.
I’ll tell you within a few days whether it’s an architecture problem and whether I’m the right person for it. Sometimes the answer is no. That’s a faster favor than a discovery call.
If the shape feels right, the rest is a conversation.