I’ll start and end this article with two recent stories (names changed to protect the awesome).
Story 1: Two Weeks vs Ten Minutes
Ben is a non-technical colleague that I respect, and I was livid that he was being taken advantage of.
His affordable, ~$30/hour dev was quoting him two weeks to deliver a feature that (I kid you not) was basically a single API query and exporting it into an HTML table.
I sat dumbfounded. I called bullshit.
This was May 2026, 6 months after the “holy shit” moment when Opus 4.5 turned the coding world upside down.
I asked Ben to let me screen-share while I fired up Claude Code. I then verbally brain-dumped a prompt.
“Make a dummy database from industry X. Then create a single HTML page. Then have that page call said database and convert it into an HTML table.”
We went back to talking. And somewhere between two and ten minutes later, the demo loaded on screen.
I turned to Ben. “So you’re telling me your developer with Claude Code is going to take two weeks to do that?”
This lunch meeting was the turning point for how his startup operated. Within days, they hired a new, AI-first developer to upskill the team. Within weeks, a client demo deadline that felt impossible was doable.
It took a technical, Forward Deployed Executive (FDX) to get the Ben over the hump.
Why I Hated the “Forward Deployed Engineer” Title
It’s ok to admit when you were wrong, and I was dead wrong.
The first time I heard about “forward deployed engineers” (FDEs), I thought it was yet another stupid marketing trick.
Isn’t this just software services?
Isn’t this just staff augmentation?
But it was fucking genius, because it solved a problem as old as time.
The Gap
In The Success Principles, mega-author Jack Canfield repeats the mantra of “closing the gap between where you are and where you want to be.”
The self-help, professional consulting, and executive coaching industries are built around this premise of gap closing.
And this gap is MASSIVE in the AI space right now. A few research examples (thank you, Perplexity!):
95% of enterprise GenAI pilots produced no measurable return. MIT’s Project NANDA looked at 300+ deployments, 52 case studies, and 153 leadership surveys; only 5% generated real financial value.
56% of CEOs saw neither higher revenue nor lower costs from AI in the past 12 months. Only 12% got both. That’s PwC asking 4,454 CEOs across 95 countries.
72% of CEOs now call themselves the primary AI decision-maker at their company, roughly double a year ago. Nearly all of them plan to keep funding AI whether or not they can prove it works.
Companies WANT to be at the bleeding edge, and yet most are feeling further and further behind… and their attempts at catching up just aren’t working.
Closing The Gap with FDEs
Palantir recognized the AI gap early. They knew the fastest way to close the gap, get customers to a win, and become indispensable for their clients was to embed the very same elite engineers that built the tech alongside their customers’ teams. Even better, these FDEs would keep identifying more opportunities to expand and grow the accounts, creating a virtuous loop.
So they coined a new job title and went all in on this approach.
I would argue that while the technique wasn’t new (it’s basically staff aug), the mindset was very different.
So how did FDE’s solve this?
Buying is Easy; Actualizing is Hard
Enterprise companies know that change is hard.
This is why sales cycles are typically 18-24 months, and most will not bring on more than 1-2 new tools per cycle.
You have to deal with things like…
Data migrations
Tooling integrations
Code refactors
Hiring new people
Training existing people
Capacity management
Comms (internal, external)
Coordination (internal, external)
Board politics
C-suite buy-in
Budgets
Mercury Retrograde (just checking if you’re still reading)
Oh, and ultimately achieve the goal or ROI you set out to achieve. Otherwise, WTF did you start this in the first place?!?
With this laundry list of challenges and headwinds, it’s no wonder why there’s a >90% failure rate for enterprise AI initiatives.
Add to that that the fact that the world is changing more rapidly than ever, sometimes on a quarterly, monthly, or weekly basis.
But all hope is not lost.
Solutions! Embedding Engineers AND Executives
The goal is to close the gap, but how?
FDEs were a critical FIRST part of that solution. They brought…
Proximity. They sit inside the customer’s team and see the actual workflow instead of the sanitized version that showed up in the RFP.
Speed. Something working in weeks (or days… or hours) rather than quarters, which breaks the pilot purgatory that produced that 95% number.
Ownership of the outcome. They’re measured on whether the customer wins, not on closing the sale or watching usage metrics through a dashboard.
But IMHO, that wasn’t enough. Yes, this addressed the technical gaps. These SEAL Team Six-type engineers can largely address business objectives, but they’re still often too far away from the key decision makers that are critical for decisions and culture.
Two citations that emphasize this.
Dan Shipper of Every once said one of the strongest leading indicators of AI adoption is whether a CEO is an avid user of AI (and usually sharing the wins to the team). I’ve witnessed this juxtaposition personally.
PwC came at the same idea from the other direction. They found that CEOs whose organizations built strong AI foundations are 3x more likely to report meaningful financial returns. The difference between the 12% who win and the 56% who don’t comes down to foundations and embedding.
The last part is critical. You need an AI foundation, which requires AI champions at the top. And since companies can’t just rapidly hire and replace the entire C-suite in a month, they need outside help… they need a Forward Deployed Executives to help level them all up.
What exactly does an FDX look like? More on that in a minute.
Remember the story of Ben? He didn’t even know he needed an FDE-like resource! He first needed an FDX, an AI-first founder type that could identify the problem and get him over the mental/emotional hurdles that prevented him from making it a priority to hire an AI-first engineer.
The AI Bottleneck is Humans
The smart folks at Anthropic (I’m looking at you, Dario) told us that AI was going to start replacing 50% of white-collar work starting by 2030.
Whoops. It doesn’t look like it’s happening… yet.
We may get their eventually, but right now the bottleneck isn’t the models, the products, the agents, or the token costs.
The gap is how fast humans can successfully absorb AI practices, products, and methods into their companies.
And that’s just the mechanics. There’s also the psychology. Here’s a direct quote from an executive I work with.
“You’re just my security blanket, so I want you here.”
AI is scary and risky. Even if someone hands you a perfect answer, they may not take it unless they trust you.
Why? The rate of change and the potential risks of AI mistakes are so high that they create so much fear, uncertainty, and doubt (FUD). It’s so massive these days that even seasoned executives who have weathered the 2008 market crash and the 2020 COVID lockdowns are affected.
Hell, there’s an executive that I’ve heard about through a friend of a friend. He’s been suffering from regular panic attacks, even throwing up multiple times a week because he’s so afraid of falling behind. I’ve heard others in my direct network say they feel the same pressure.
These execs don’t need another LinkedIn article with 5 new tips. They need a trusted confident to describe their challenges and, most importantly, help them close the gap.
They need someone with both technical AND leadership experience that can help them navigate this chaos and complexity.
What Makes a Good FDX
History doesn’t repeat, but it does rhyme.
A decade ago, one of the hottest hires on the market was a “solutions architect.” This mythical hire had skills and experiences across 3 key areas.
Technical (i.e. deep understanding of architecture, best practices, etc)
Management (i.e. costs, timelines, team makeup, etc)
Entrepreneurial (i.e. business needs, ROI, market opportunities)
You may recognize these factors from the bestselling book The E-Myth.
And rare was the person who could do all three. They were outstanding hires.
Now we’re in an era where the gap requires all three of these plus a few more.
Adaptable (i.e. able to react rapidly, with high tolerance for change)
AI-first (i.e. approaching everything from an AI-first mentality)
DIY, DWY, DFY (i.e. the ability to build, teach, and delegate these skills)
I’ve been doing AI upskilling now for over a year, and you’d be amazed at the number of executives who want and need a coach to sit in the trenches with them (to test, tinker, optimize, vent, etc.). They need someone that’s like them enough to translate into their words and needs as well as technical enough to knock down walls with and for them.
Can I Help?
Ok I didn’t mean to make this a pitch, but why not? Here’s a bit about me:
MIT BS + PhD.
18 years of startup experience (6x co-founder)
Enterprise, SMB, and security/compliance experience.
Techstars Mentor x5 cohorts.
Coached hundreds of people.
12 months of AI-first consulting, training, up-skilling.
Wrote a book, Save Your Startup, to help founders with actionable playbooks.
Enough about me. What matters most is an honest question to. How can I help?
I’d be happy to riff on any ideas. DM me on LinkedIn, X, or Substack.
Now back to the article close…
Story 2: “My CTO is an OpenClaw”
Tyler is a non-technical colleague that I respect, and I was thrilled he trusted me with such an ambitious bet.
I had been sharing AI news with him for weeks/month. Then on Sunday, January 25th, 2026, he asked, “Thoughts on clawdbot?” (later rebranded to OpenClaw). I basically shared that I bought a Mac mini and was about to go all-in. I said it would be as pivotal as Opus 4.5, and the world would understand this in under a month.
Tyler didn’t just say “cool” and move on. Within one week, he had two OpenClaws running his company, one as the CTO. By week two, he was already shipping 3-5x and reduced his team size by half. He’s the CEO of a small, and aggressively AI-first startup.
Similar to Ben, an FDE wasn’t the bottleneck (at first). The first step was another colleague, but an aggressively AI-pilled one, to push him to take a swing. And swing he did. Now, the student became the teacher. Tyler is now so far ahead of me it’s not funny. And that’s awesome. We continue to push each other.
The Billion Dollar Opportunity
I made a bold claim in the title. Is this really a billion-dollar (or more) opportunity?
Absolutely.
But did you catch that I used the word “unblock” vs “unlock” in the title? Why?
First, “unlock” is an overused AI-slop word… abused like my previous favorite punctuation, the em-dash.
It’s also the wrong metaphor. What’s sitting in the way is more like blockage of our own creation. It’s human, it’s emotional, and clearing it is unglamorous work that somebody has to physically show up and do.
US companies are easily spending 100’s of billions of dollars in aggregate to adopt AI as fast as possible.
And again, >90% of these things fail to achieve the ROI… yet.
FDEs were a key part of that solution for forward-thinking companies like Palantir. However, industry-wide, it’s not enough.
If we see enough FDX’s help scope, scale, and deliver better results and move these percentages higher, we’re going to see higher percentages of AI adoption with real, measurable ROIs. This will justify additional investment. This will accelerate a virtuous cycle.
A billion-dollar opportunity is awesome, but that’s just the beginning. The bigger story is the career opportunity for executives worried about AI replacing them (no need to have panic attacks when the amount of inbound demand is rising). Maybe AI will replace us all eventually, but not today in 2026. Today there’s a gold rush of opportunity because there are gaps everywhere, and businesses are just looking for people to help close these gaps.
Maybe that’s you.
I know it’s certainly me.
Which is why I’m so excited about where my career is headed leading into 2027.
I hope you’re feeling more hope than anxiety.
Until next time ✌️
PS. Wrote By Me
There’s a trend these days of Substack articles being 90-100% AI-generated with little to no human editing. That’s a shame. This article is the opposite. I rewrote this draft 4 times by hand before asking AI for assistance, pulling in some stats and research. Then I used Grammarly for spelling, grammar, and brevity suggestions. Then I rewrote sections again because I’m determined to focus and followthrough.
I reject many of Grammarly’s suggestions because I like an occasional cuss word. I also love run-on sentences, ellipses, and other writing no-nos. In the age of AI perfection, I prefer to be proudly imperfect. I hope you’d consider approaching your writing the same way.
Thought provoking article Rick!