You’ve probably noticed the latest flex in the tech world. Founders aren’t calling themselves software developers anymore; they’re “Forward Deployed Engineers” or FDEs. It sounds elite. It sounds like you’re on the front lines of the AI revolution. But let’s be brutally honest: most of you aren’t doing FDE work. You’re just running a traditional outsourcing grind and putting a trendy label on it.
The term was popularized by Palantir, and the original concept was brilliant. Palantir talked about building “gravel roads” and “highways.” The idea was to go on-site with a client, build a messy, temporary, custom solution (the gravel road) to get them moving immediately. But the real goal wasn’t the gravel road—it was the highway. You were supposed to take that custom experience, extract the repeating patterns, and build a scalable, standard product that you could sell to everyone else.
Project revenue is the most dangerous drug in tech. It feels like survival, but it’s actually a slow death.
But that’s not what’s happening right now. The gravel road has become the destination. I talk to founders all the time who are exhausted, burnt out, and operating on razor-thin margins. Why? Because they lack the strategic discipline to say no to custom work. They take every check that comes through the door, building bespoke solutions that can never be scaled or reused. They’ve traded their product moat for low-margin, unscalable labor.
Here is the dirty secret nobody tells you about these clients: they aren’t just outsourcing AI development. They are outsourcing uncertainty.
You aren’t being paid to engineer solutions. You’re being paid to do the thinking your client is too paralyzed to do themselves.
Most clients haven’t even defined their problem. They just know “AI” is a buzzword and they need it. They hand you a pile of vague requirements and pay you to figure out what they actually want. You’re doing their internal strategy, not engineering. And when the engagement ends, you have nothing to show for it except a tired team and a one-off script that will rot in a codebase.
True FDE isn’t a job title; it’s a business model. It’s the discipline to look at a project and ask: “Will this result in a product I can sell to the next ten clients?” If the answer is no, you walk away. Even if the cash is sitting on the table. Because the moment you accept a custom project that can’t be standardized, you’ve stopped being a product company. You’ve become a services shop.
A gravel road is only valuable if it leads to a highway. If you stop paving, you’re just stuck in the mud.
If you’re a tech founder or product leader, you need to wake up. Stop mistaking project revenue for Product-Market Fit. The goal isn’t to deliver a project and get paid. The goal is to solve a common pain point, achieve a business result, and then replicate that solution at a lower cost for the next customer. If the next customer requires you to start from scratch, you haven’t found PMF—you’ve just found a new way to bill hourly.
Drop the FDE title if you’re just doing implementation. Own the fact that you’re building a product, and act like it. The companies that win the next decade won’t be the ones with the most billable hours. They’ll be the ones who had the guts to stop doing custom work long enough to build the highway.
FAQ
Q: Isn't taking on custom work necessary for early-stage startups to survive?
A: Survival is a valid short-term need, but if every project requires a bespoke build, you're building a services agency, not a SaaS startup. You're trading long-term equity for short-term payroll.
Q: How do I know if a custom project is worth taking as an FDE?
A: The test is simple: Can the solution be replicated for the next three clients with minimal engineering effort? If the answer is no, it's pure outsourcing, not product discovery.
Q: What if the client is willing to pay a massive premium for exclusivity?
A: Take the money, but don't lie to yourself. You're building a bespoke asset for them, not your company's moat. Use the cash to fund your actual standard product, or you'll be stuck on the hamster wheel forever.