Hey everyone, Neil here. You're reading High-Signal Hiring. Hiring systems from 20+ years of global recruitment experience and 500+ technical hires. Zero noise and instantly actionable.
Last issue we broke "AI Engineer" into five distinct profiles and showed why benchmarking off Levels.fyi without a discount torches your runway.
This week, the role hiding inside most of those JDs. The one most founders need without knowing what it’s called.
It's called Forward Deployed Engineer (FDE for short). Postings grew 1,165% year-on-year through October 2025. Salesforce committed to a thousand of them! OpenAI, Anthropic, Palantir, Databricks, Ramp, Rippling, and Intercom are all hiring them. The Pragmatic Engineer recently called it the hottest job in tech.
You'll learn what an FDE does, whether you're ready to hire one yet, the comp band by stage, and the interview round to add to the Issue #19 loop.
Not a subscriber yet? Sign up here
| What an FDE does
They scope a customer integration in week one, ship it in week three, fix the data-quality problems no one anticipated in week four. By week six the customer is using it daily and there's a pattern your product team can generalise.
The skill set is a T-shape. Deep enough to ship Python or TypeScript against AWS, GCP, vector databases, and agent orchestration tooling. Broad enough to sit across from a customer's CTO and translate a constraint into a design decision in real time.
Communication is not a soft skill in this role. It's the job. Half the value is the build, half is the conversation that gets you to the right thing to build.
| Are you ready for one?
Three tests before you spend on this hire:
1️⃣ Do you have a paying customer with a non-trivial integration to ship?
2️⃣ Does your roadmap require customer-by-customer customisation, not just feature work?
3️⃣ Are you spending 40%+ of your founder time in customer-facing build conversations?
Yes to two of three, you're FDE-ready. Fewer than two, then you’re probably jumping the gun. An FDE without a customer runs out of work in week three and quits by month three. Hire a founding engineer instead and come back when you have customer pull.
| Why most founders write the wrong JD
Open ten "Senior AI Engineer" job descriptions at seed and Series A startups shipping AI features. Most describe FDE work. Customer-facing AI feature delivery, production deployment with real users, integration with messy enterprise data. A handful describe ML or Applied AI work. Almost none describe AI Research Engineer work, even though the comp band quoted is anchored there.
The mismatch costs founders twice. The JD attracts the wrong candidates, researchers who don't want to talk to customers and SWEs who don't want to debug a customer's auth setup. Then the comp band is wrong, either over-paying for a research profile you don't need or under-paying for the customer-shaped operator you do.
On the SEO trade-off. "Forward Deployed Engineer" gets fewer searches than "Senior AI Engineer." Dual-list the title. FDE language in the body. You keep the inbound and filter on the way in.
| The comp band by stage
FDE comp doesn't sit where the AI engineer headlines suggest. The band stretches more by stage than by skill.
Median across 1,000+ FDE postings: $173,816. Palantir's FDE median: $215K total comp. OpenAI runs $350-550K for mid-to-senior, staff-level clearing $630K. The open market lands $200-400K total.
By your stage (US):
Pre-seed / seed. $130-160K base, plus 0.5-1.5% equity. The equity is your differentiator, not the cash. Can't get to $130K base? Try a 3-month contract-to-hire at a senior consultant rate and convert if it works.
Series A. $170-200K base, ~ $220-300K total comp.
Series B+. $200-260K base, ~ $280-400K total comp.
London and Berlin take a 30-40% discount on US comp before currency. SF and NYC sit at the top of the pile.
The premium over a senior SWE is real. The combination of customer face time and shipping production code is rare.
| How the Depth round changes for an FDE
The four-stage, three-interview loop from Issue #19 is the structural anchor. Founder screen, Mission and Depth, Alignment, Reference and close. For an FDE, you don't add a round. You change what Depth tests.
In Interview 2, run a Solution Design problem inside the Depth section. Bring an open-ended customer scenario the candidate would face in their first 90 days. "We want to land Acme as a design partner. They have unstructured PDF data and want our agent to operate over it. They have one engineer and three weeks. How would you scope this?"
Watch for four signals:
They ask three or four sharp questions about the customer before talking technology.
They sketch a phased plan with what ships in week one versus week four.
They flag what's likely to break and how they'd handle each break.
They talk about what they'd negotiate down with the customer to make the timeline work.
Two showing up clearly, you're talking to an FDE. The signal isn't a checklist score. It's whether they think customer-out or technology-out. A research-shaped candidate goes straight to architecture. A senior SWE asks for a tighter spec. Neither is your hire.
| Where to find them
👉 The aliases to search for:
"Forward Deployed Engineer" OR "Solutions Engineer" OR "Implementation Engineer."
On LinkedIn, layer in keywords like "shipped to production," "customer integration," and your stack. Filter to 3-7 years experience. Send a tailored note that names the customer scenario you're hiring against.
👉 The CV shape to chase:
Engineers at consultancies who left to ship product. Solutions engineers who graduated to writing the code. Founding engineers whose startup didn't make it but shipped customer-facing AI in the last 18 months. Ex-Palantir is bid up.
| What to do this week
Three things.
Run the readiness test: If you don't pass two of three, stop here. Hire a founding engineer or generalist and come back to this issue when you have customer pull.
Open your current "AI Engineer" JD: Dual-list the title. Rewrite the body around "ship to a real customer in three weeks" outcomes. Reset the comp band to your stage.
Run Depth as Solution Design: Inside Interview 2, swap the standard Depth questions for a customer scenario the hire would face in their first 90 days. Watch the four signals. Make the call on feel, not score.
The market has named this role, priced it by stage, and built the interview shape. Use the framing and you hire in 4-6 weeks. Stay on "Senior AI Engineer" autopilot and you pay more, take longer, and hire a researcher who won't talk to a customer.
For a deeper read on the FDE profile, The Pragmatic Engineer's piece is the best single source.
Next week: the offer letter footnote that quietly kills 30% of close conversations.
Cheers
Neil