Paste your requirements.
Get a ranked shortlist.

Here’s my read before I start searching:

• This is a deployment-builder hire, not a classic pre-sales SE search — the closest matches will be people who still code, work inside customer environments, and turn messy enterprise problems into production AI systems.
• I see three real candidate lanes — career FDEs / forward-deployed builders, applied AI engineers with consulting or post-sales delivery depth, and technical founders or early startup engineers who’ve done customer-facing implementation. I’d start with the first lane unless you want more 0→1 founder energy than title fidelity.
• The hardest trade-off is frontier AI depth vs enterprise delivery polish — the market often splits those between AI-native builders and cloud/consulting operators. Your preference there will change both target companies and how I score adjacent profiles.
• “Founding FDE” likely means high-agency ICs, not heavyweight leaders — I’ll bias toward hands-on builders who can shape the motion, unless you want someone meaningfully more senior than the title suggests.

Here’s what I’m seeing in the market:

• London is a real target pool, not just a fallback geography — Anthropic, OpenAI, Palantir, Faculty, Arva AI, and Mistral all show applied AI or forward-deployed talent in-market there. That means I can stay tight on London rather than widening to Europe immediately.
• Palantir-style forward-deployed builders are still the cleanest translation — that ecosystem consistently produces engineers who embed with customers, ship in production, and operate well under ambiguity. I’m treating that lane as a benchmark, even though you want founder-heavy coverage first.
• The founder lane is viable, but only if I stay strict on hands-on LLM evidence — plenty of startup founders look strategic; far fewer show production agents, evaluations, Python builds, or enterprise deployments. I’ll screen that lane hard for real build depth.
• The best adjacent pool is applied AI delivery, not generic solutions engineering — Databricks-, OpenAI-, and Faculty-style profiles with RAG, agent orchestration, and evaluation work are much closer than classic pre-sales architects. I’m biasing toward builders who happened to be customer-facing, not sellers who learned AI vocabulary.

Founder-heavy profiles will only be worth it if they show real build depth, so I’m testing that lane against cleaner FDE and applied-AI delivery pools rather than assuming the founder title itself is enough.

The early read is clear: Palantir/OpenAI-style forward-deployed profiles are the most reliable lane, while founder backgrounds only work when the profile still reads like a hands-on builder rather than an executive.

I’ve got enough signal now: the London pool is real, but the market splits between clean forward-deployed operators and founder/builders who only sometimes translate into customer-embedded delivery.

The market is splitting the way I’d expect: the cleanest fits are Palantir/OpenAI-style forward-deployed operators, while the founder lane only works when the profile still reads like a builder, not a startup executive.

Shortlist Summary — Forward Deployed Engineer, Applied AI, Anthropic

36 candidates sourced

🟢 Tier 1 — Contact First These are the profiles I’d put at the top of outreach because they either already do the closest version of this job, or they give you the founder-style builder angle without losing hands-on AI delivery credibility.

🔵 Tier 2 — The Rest of the Pool (30 candidates)
The remaining shortlist breaks into three usable groups.
About a dozen are Palantir/OpenAI/Databricks-style forward-deployed or deployment-strategist profiles — the cleanest title translation, though some skew a bit junior or a bit lead-heavy.
Another meaningful cluster is founder / founding-engineer talent with real agent, eval, and LLM build depth — stronger on 0→1 energy than on repeat enterprise delivery.
The rest are applied AI engineers and hands-on consultants in London who show solid RAG / GenAI shipping evidence, but need screening on whether they were actually embedded with customers or mostly building internal tools.
The best outreach angle across this group is “come build Anthropic’s field engineering motion from the beginning,” especially for people currently doing adjacent work inside OpenAI, Palantir, or independent AI delivery roles.
One especially notable reserve profile is [Kai Kowalski](https://www.linkedin.com/in/slate-connors-72b1d6 who bridges OpenAI and Palantir well but reads a bit more solutions-architecture than pure FDE.

What I learned

• Palantir is still the cleanest factory — that ecosystem produces the most obvious customer-embedded operators in London, and it keeps outperforming broader “applied AI” phrasing.
• Founder fit is narrower than it sounds — the good founder profiles are very good, but most founder results drift too strategic, too senior, or too light on enterprise deployment ownership.
• London is workable without widening yet — there’s enough density across Palantir, OpenAI, founder-led AI companies, and independent builders that I don’t think geography is the bottleneck right now.
• The biggest interview question won’t be visible on LinkedIn — for many of these candidates, the real screen is how much white-glove customer deployment they’ve personally owned versus advising, architecting, or managing around it.

Take a look through the shortlist and tell me which lane feels most right to you. If you want, I can do the next pass either tighter on Palantir/OpenAI-style operators or more aggressively into founder/builder profiles with strong MCP-agent depth.

I put together a shareable market brief for the hiring manager that explains:

• how the London market splits across the main candidate lanes
• where the strongest talent clusters sit
• what backgrounds translate best to Anthropic
• what looks close but usually isn’t
• the practical implications for calibrating interviews and expectations

If you want, I can also turn it into a shorter one-page version with a more executive tone.

Tightened it into a shorter version. If you want, I can make it even more exec-style — closer to a half-page brief.