Plank Research · live census · 2026-06-12

The FDE job market, mapped

We collected 991 live forward-deployed engineer postings across 569 companies, from company ATS boards (Greenhouse, Lever, Ashby), a large ATS-indexing job aggregator, and sweeps of Y Combinator, Wellfound, LinkedIn, Indeed, and VC job boards. Title-matched and de-duplicated to one row per real role, each individually linked. A posting is one job description, not one hire, and a single Palantir or frontier-lab req can staff many, so read every count here as a floor. That is roughly half the strict global FDE-titled market (~1,000-1,400), weighted to the highest-signal employers. How we get to that number, and why broader definitions reach 20,000+, is the next section. The charts synthesize what we verified; the table lets you drill down to any single posting.

live FDE postings collected, every one individually linked below
991
live FDE postings collected, every one individually linked below
companies hiring FDEs, from frontier labs to seed-stage YC startups
569
companies hiring FDEs, from frontier labs to seed-stage YC startups
median posted base midpoint, across the 260 postings that disclose pay
$185K
median posted base midpoint, across the 260 postings that disclose pay
of openings sit at frontier labs and Palantir, the enterprise-facing core
11%
of openings sit at frontier labs and Palantir, the enterprise-facing core

Six things the data makes clear

Every figure links to the chart it came from. Hover to see the source.

  1. 01

    A company’s first forward-deployed-engineer req is the earliest public signal it is shifting from product-led to enterprise revenue, a leading indicator now visible across 569 companies and 22 verticals after 729% year-over-year posting growth.

  2. 02

    The role is more global than its reputation: two-thirds of the openings sit in the United States and a third run international, yet US demand still clusters on the coasts at a median 22% above the heartland, pricing a travel-defined job as if nearness to capital, not to the customer, set the wage.

  3. 03

    By a Series B the median base ($235K, within $11K of a frontier lab’s base) all but matches the labs, but base is where the resemblance ends: equity is 55 to 70% of a frontier-lab package, pushing total compensation there past $700K, so the two look alike only on the one line that matters least.

  4. 04

    95% of enterprise GenAI pilots still return nothing to the P&L, and the one variable that doubles the odds of reaching production is not a better model but embedding the engineers who deploy it.

  5. 05

    An employer’s vertical shifts a forward-deployed engineer’s pay more than seniority does: $250K at a foundation-model lab versus $148K in industrial AI is a $102K swing, wider than the $71K gap from mid-level to head of the function.

  6. 06

    The frontier labs now pitch operating companies forward-deployed engineers who deploy their own stack, but since deployment is only about 20% of an AI system’s lifetime cost, the lock-in lands in the 80% paid for later.

How many FDE jobs are there, really?

It depends entirely on where you draw the line, so here is the whole ladder, from an exhaustive sweep of every major job source. Roles literally titled “forward deployed engineer” number ~1,000-1,400 worldwide, de-duplicated. But the job itself, a customer-embedded engineer who ships in production, is also done by solutions, implementation, and field engineers; count all of them and you clear 20,000 globally. What isn’t real is the “56,000+” LinkedIn shows: that’s OR-matched word-soup, and Indeed’s own count drops ~70× the moment you restrict it to the title.

  • All customer-embedded delivery roles

    18,000-25,000

    Solutions, implementation, field, and professional-services engineers worldwide. Many do the same job an FDE does, embedding with a customer and shipping in production under a different title. The solutions-engineer family alone is roughly 18,000 to 24,000 worldwide across the major ATS aggregators. This is the definition under which “20,000+ FDE jobs” is true.

  • FDE + AI-era equivalents

    3,000-6,000

    The FDE title family plus the AI-native rebrands doing identical work: applied AI engineer, customer engineer, agent engineer, forward-deployed AI engineer. The boundary here is a judgment call, not a clean measurement.

  • All FDE-titled variants, de-duplicated

    1,500-2,000

    Every genuine FDE title (“forward deployed,” “deployed,” and “field deployed” engineer), de-duplicated, global, with roughly 900 to 1,300 in the US. Aggregator data shows most “deployed engineer” titles already are forward-deployed, so this barely exceeds the strict count.

  • Strict “forward deployed engineer,” de-duplicated

    ~1,000-1,400

    Roles literally titled forward deployed engineer, de-duplicated across hundreds of company ATS systems. The leading ATS aggregator's exact-phrase count is 1,031 live worldwide (1,391 all-time), and 650 to 885 in the US. Growing fast: Indeed indexed +729% year over year.

  • Plank census, verified & individually linked

    991

    The subset we confirmed live and linked one by one, across 569 companies, each row pointing to the company's own posting. Built by re-scanning every employer's ATS and a site-scoped sweep of Ashby, Lever, Greenhouse, Rippling, and Workable, counting distinct role-and-location postings (so a role open in several cities counts once per city). That it nearly reaches the strict-titled estimate above, built from a single aggregator's de-duplication, suggests that estimate is conservative.

Both answers are true. They ask different questions. Roles literally titled “forward deployed engineer” number roughly 1,000 to 1,400 worldwide and are climbing fast. But the job itself, a customer-embedded engineer who ships in production, is also done by solutions, implementation, and field engineers, and counting all of them clears 20,000 globally. What you won't find is tens of thousands of roles actually titled FDE. The “56,000+” on LinkedIn is OR-matched word-soup, not real demand.

Who is hiring FDEs, and where

Demand concentrates hard: Palantir and the frontier labs account for most open roles, but the long tail, growth-stage AI products hiring their first FDEs, is where the category is spreading fastest.

Top employers

Open FDE postings per company (unique live postings, not headcount). Beyond these, a long tail of 557 companies accounts for 818 more postings, mostly one or two roles each.

Plank FDE census · n=991 · 2026-06-12

By segment

Frontier labs sell deployment; Palantir invented it; growth startups are adopting it.

  • Startup / scale-up678 · 68%
  • Growth startup109 · 11%
  • Public tech & enterprise95 · 10%
  • Frontier lab62 · 6%
  • Early-stage startup43 · 4%
  • Consulting & SI4 · 0%
Plank FDE census · n=991 · 2026-06-12

By region

Posting locations normalized to region. Multi-location postings counted once at the primary location.

Plank FDE census · n=991 · 2026-06-12

By industry vertical

Each employer's primary vertical. Defense, foundation-model labs, and data/AI platforms lead the named categories; the large 'Enterprise software' bucket is the catch-all for the long tail of AI companies a name-and-title keyword pass could not place more precisely.

Plank FDE census · n=991 · 2026-06-12

Posted base-salary ranges

The 14 companies with the most disclosed-pay postings (of 260 postings disclosing pay). Bars span the lowest posted minimum to the highest posted maximum; n = postings with disclosed pay.

Plank FDE census · n=991 · 2026-06-12

By seniority

Inferred from titles. Most postings don't level the role in the title, consistent with FDE being hired as a senior-by-default, single-track role.

Plank FDE census · n=991 · 2026-06-12

What the job descriptions ask for

We parsed the full text of 674 forward-deployed engineer job descriptions and counted what they actually require. Each bar is the share of descriptions that name a given technology or skill. Two things stand out: the role runs on Python and the modern LLM stack, and what separates it from an ordinary backend job is the human part, customer-facing work, travel, and stakeholder management.

Languages

Share of analyzed job descriptions that name each, by document frequency.

Plank FDE census · n=991 · 2026-06-12

Cloud & infra

Share of analyzed job descriptions that name each, by document frequency.

Plank FDE census · n=991 · 2026-06-12

Data

Share of analyzed job descriptions that name each, by document frequency.

Plank FDE census · n=991 · 2026-06-12

AI / ML

Share of analyzed job descriptions that name each, by document frequency.

Plank FDE census · n=991 · 2026-06-12

Capabilities expected

Share of analyzed job descriptions that name each, by document frequency.

Plank FDE census · n=991 · 2026-06-12

Document frequency: share of analyzed JD bodies that mention each item. Base = ATS-fetchable descriptions (Greenhouse, Lever, Ashby, Databricks); aggregator, LinkedIn, and Indeed bodies are not machine-readable.

Cross-reference the market

Two patterns jump out of the base-pay data: base climbs with company maturity, and a coastal premium persists even for a role defined by travel. Read both with the equity caveat below, since base is the minority of a frontier-lab package.

Stage × pay

$165K → $246K

Median posted base climbs from seed-stage startups to frontier labs, a 49% spread across the maturity curve. Series B and late-stage companies cluster in between. But this is base only: the labs’ equity-led packages push total comp far past the top bar, see the note below.

Geography × pay

+21% coastal

FDE roles in coastal hubs (the Bay Area, NYC, Seattle, Boston, the DC corridor) post a median $185K base versus $153K across the heartland and Sun Belt, a premium that holds despite the role’s travel-heavy, deploy-anywhere nature.

Median posted base pay, every cut

By funding stage

Pay climbs with company maturity, from seed to frontier lab.

Plank FDE census · n=991 · 2026-06-12

By geography

Coastal hubs pay a persistent premium over the heartland.

Plank FDE census · n=991 · 2026-06-12

By industry

Where the disclosed-pay sample supports a median.

Plank FDE census · n=991 · 2026-06-12

By seniority

Inferred from the posting title.

Plank FDE census · n=991 · 2026-06-12

Median of posted base-salary midpoints. Only buckets with at least 5 disclosed-pay postings are shown, so a few thin categories are omitted.

Base is not the offer

What posted base leaves out: equity

Every salary figure above is posted base, the only pay companies disclose at scale (just 26% of postings disclose even that). But at the labs and platforms base is the minority of the package. Posted base pay tells less than half the story: at frontier AI labs equity, not salary, is now the majority of total comp, lifting senior-IC packages from a ~$300K base to $700K-$900K+ all-in.

Total comp: base + equity

Median engineering total compensation at a mid-senior IC level, base plus annual equity, in thousands. FDE and solutions roles map to this software-engineer ladder. Source: Levels.fyi (US), June 2026.

Plank FDE census · n=991 · 2026-06-12

Equity as a share of total comp

The more frontier the employer, the more of the offer is stock, not salary.

  • Frontier AI labs44-70%$560-1000K total
  • Late-stage / public tech27-65%$211-663K total
  • Growth-stage AI startups33-60%$300-640K total
  • Early-stage / seed startups20-50%$130-300K total

OpenAI Profit Participation Units (PPUs)

A PPU (Profit Participation Unit) was OpenAI's primary equity-style compensation instrument through 2024-2025: unlike RSUs or options, it conferred no ownership, no voting rights, and no strike price, just a contractual claim on a slice of OpenAI's future profits, granted at no cost to fit the company's then-"capped-profit" structure. PPUs vested 25% per year over four years with no cliff, plus a roughly 2-year lock-up, and had no public market, value was set by the most recent price investors paid in company-run secondary tenders (most prominently the ~$6.6B October 2025 secondary at a ~$500B valuation), with legacy upside capped at 10x the original grant value. A representative mid/senior engineer offer was ~$300K base plus a ~$2M PPU grant vesting over four years (~$500K/yr), so equity ran roughly 1.5-3x base annually, making it ~60-75% of total comp. Note the timeline: after OpenAI's October 28, 2025 conversion to a Public Benefit Corporation, outstanding PPUs convert to conventional shares (removing the 10x cap) and reporting indicates new grants moved to standard RSUs from around January 2026, so "PPU" describes the 2024-2025 era while 2026 hires increasingly receive RSUs.

Anthropic pays in RSUs on private stock (4-year vest, 25% in year 1 then ~2.08%/month, effectively a 1-year cliff) with no cash bonus for engineering, comp is base plus equity only, where equity runs ~44% of total for Senior SWE ($563K) and rises to ~58% at Lead SWE ($785K). Because those RSUs are on illiquid private stock, realized value depends on tender/IPO timing against a fast-climbing 2026 valuation (reported into the hundreds of billions, with a confidential IPO filing), pairing significant upside with a real illiquidity window.

So read the frontier-lab and foundation-model pay above as a floor on base, not the offer: the real gap from an early-stage startup to a lab is measured in equity, not the $81K of base between them.

Total-comp sources: Levels.fyi · Levels.fyi · Levels.fyi · Perspective 2026 FDE report · Levels.fyi · CNBC

Every posting

The full census: search it, filter it, and click through to any live posting. Or download the dataset (JSON, CC BY 4.0).

991 postings · showing first 50

CompanyRoleLocationIndustryStageSalarySource
11xForward Deployed EngineerUnspecifiedEnterprise softwareSeed / stealthn/aCompany site
Abacus InsightsSenior Sustaining & Forward Deployed EngineerRemote USEnterprise softwarePrivate / undisclosedn/aGreenhouse
AccentureForward Deployed Engineer ManagerUnspecifiedConsulting & staffingPrivate / undisclosed$87K-$294KWorkday
Accenture Federal ServicesForward Deployed Engineer - Palantir FoundryWashington, DCConsulting & staffingSeed / stealthn/aGreenhouse
Accenture Federal ServicesForward Deployed Engineer, Palantir Foundry - Honolulu, HIHonolulu, HIConsulting & staffingSeed / stealthn/aGreenhouse
AccreteDirector, Forward Deployed EngineeringNew York, NYEnterprise softwareSeed / stealth$240K-$280KBreezy
AccreteSenior Forward Deployed EngineerNew York, NYEnterprise softwareSeed / stealth$160K-$200KBreezy
AcrelyForward Deployed EngineerUnspecifiedEnterprise softwareSeed / stealth$90K-$120KYC
AdaptionFounding Forward Deployed Machine Learning EngineerSan FranciscoEnterprise softwarePrivate / undisclosedn/aAshby
Adaptive MLForward Deployed AI EngineerNew York OfficeEnterprise softwareSeed / stealthn/aAshby
AddeparForward Deployed EngineerRemote, USAEnterprise softwareSeed / stealthn/aGreenhouse
AdobeForward-Deployed AI EngineerUnspecifiedEnterprise softwarePrivate / undisclosed$126K-$234KWorkday
AdonisSenior Forward Deployed EngineerNew York CityHealthcare & pharma AIPrivate / undisclosed$190K-$210KAshby
AI FundForward Deployed EngineerMountain View, CAEnterprise softwareSeed / stealthn/aLever
AI FundForward Deployed Engineer (Ayama)San Francisco Bay Area, RemoteEnterprise softwareSeed / stealthn/aLever
Aimpoint DigitalForward Deployed AI Engineer - Claude 2026 - USAtlanta, GAEnterprise softwarePrivate / undisclosedn/aBreezy
Aimpoint DigitalLead Forward Deployed Engineer, Databricks 2026- US, UKAtlanta, GAEnterprise softwarePrivate / undisclosedn/aBreezy
AircallDirector, Forward Deployed Engineering & AI GTM LeadSeattle OfficeVoice & video AIPrivate / undisclosedn/aLever
AircallDirector, Forward Deployed Engineering & AI GTM LeadSan Francisco OfficeVoice & video AIPrivate / undisclosedn/aLever
AircallDirector, Forward Deployed Engineering & AI GTM LeadNew York OfficeVoice & video AIPrivate / undisclosedn/aLever
AircallForward Deployed Engineer - AI Solutions EngineeringNew York OfficeVoice & video AIPrivate / undisclosedn/aLever
AircallForward Deployed Engineer - AI Solutions EngineeringSeattle OfficeVoice & video AIPrivate / undisclosedn/aLever
Airkit (Salesforce)Agentforce Forward Deployed Engineer, B2C CommerceManila / New York / RemoteCustomer-service agentsPublic companyn/aGetro · Accel
Airkit (Salesforce)Forward Deployed Engineering LeadChicago / RemoteCustomer-service agentsPublic companyn/aGetro · Accel
AirweaveFounding Forward Deployed EngineerUnspecifiedEnterprise softwareSeed / stealthn/aYC
AIX VenturesForward Deployed EngineerUnspecifiedConsulting & staffingConsulting / SIn/aGetro · Khosla
AkkaForward Deployed Engineer (North America)New York, NYDeveloper infrastructureSeries C+ / laten/aLinkedIn
AlationPrincipal Forward Deployed EngineerRedwood City, California, United StatesData & AI platformsPrivate / undisclosed$213K-$287KWorkday
AlbacoreForward Deployed EngineerPhiladelphia, Pennsylvania, United StatesEnterprise softwarePrivate / undisclosedn/aGem
AledadeSenior Software Engineer I- (Forward Deployed AI)Remote, United StatesEnterprise softwarePrivate / undisclosedn/aLever
AllanaSenior Forward Deployed EngineerUnspecifiedEnterprise softwareSeed / stealthn/aCompany site
Allegis GroupDeveloper (Forward Deployed Engineer)UnspecifiedDeveloper infrastructureSeed / stealthn/aCompany site
Altus GroupForward Deployed EngineerHouston, TX (hybrid)Data & AI platformsPublic company$185K-$200KIndeed
AmariFounding Forward Deployed EngineerSan FranciscoEnterprise softwarePrivate / undisclosed$180K-$260KAshby
AmazonForward Deployed AI Integrator, Field EngineeringUnspecifiedEnterprise softwarePrivate / undisclosed$127K-$172KAmazon
Amazon (AWS ProServe)Forward Deployed Engineer, WWPS ProServeArlington / Herndon / Denver / SeattleAI infrastructurePublic company$131K-$178KCompany site
AmplitudeEngineering: Senior Customer Forward Deployed EngineerSan Francisco, CAEnterprise softwarePrivate / undisclosedn/aGreenhouse
AndersenManager - Foundry/AIP Forward Deployed EngineerSan Francisco, California, United StatesEnterprise softwarePrivate / undisclosed$196K-$266KCompany site
AndurilForward Deployed Engineer, Mission AutonomyCosta Mesa, CADefense & data platformsSeries C+ / late$113K-$155KGetro · 8VC
Anduril IndustriesForward Deployed Engineer, Air Defense (FDE)Costa Mesa, California, United StatesDefense & data platformsPrivate / undisclosedn/aGreenhouse
Anduril IndustriesForward Deployed Engineer, Mission AutonomyCosta Mesa, California, United StatesDefense & data platformsPrivate / undisclosed$113K-$155KGreenhouse
AnnealForward Deployed EngineerUnspecifiedIndustrial & logistics AISeed / stealthn/aGetro · Amazon Catalytic
AnthropicForward Deployed Engineer, Applied AIBoston, MAFoundation modelsFrontier labn/aGreenhouse
AnthropicHead of Forward Deployed EngineeringSan Francisco, CAFoundation modelsFrontier labn/aGetro · Menlo
Applied ComputeForward Deployed EngineerSan FranciscoEnterprise softwarePrivate / undisclosedn/aAshby
AravoltaForward Deployed EngineerSan Francisco / Remote (US)Industrial & logistics AISeed / stealth$60K-$240KYC
ArcadeForward Deployed EngineerSan Francisco, CAEnterprise softwarePrivate / undisclosedn/aAshby
ArcadeForward Deployed EngineerAustin, TXEnterprise softwarePrivate / undisclosedn/aAshby
Archer SystemsSenior Forward Deployed AI EngineerHouston, Texas, United StatesEnterprise softwarePrivate / undisclosedn/aUKG
Arize AIForward Deployed AI Engineer, EMEARemote (EMEA)Data labeling & evalsSeries C+ / laten/aGreenhouse

Methodology & caveats

Collection. Postings were observed on 2026-06-12. Collection ran in three passes: (1) public ATS APIs (Greenhouse boards, Lever postings, and Ashby job boards) for companies known or suspected to hire FDEs, matching titles against forward deployed / FDE; (2) a large job-board aggregator that indexes postings directly from company ATS systems, paginated through the full result set for the quoted phrase, with every aggregator row resolved back to and linked at the company’s own posting; (3) manual sweeps of Y Combinator’s Work at a Startup, Wellfound, LinkedIn, Indeed, Google/Salesforce/IBM/AWS careers, and Getro-powered VC portfolio boards; and (4) a re-scan of every ATS board surfaced by the first three passes, pulling each employer’s full board to catch sibling forward-deployed roles (other cities, seniorities) we had not yet linked. Everything was de-duplicated by normalized URL and by company-plus-title-plus-location, so an identical role reposted in the same city counts once while genuinely distinct cities stay distinct; ATS and verified-board rows take precedence over aggregator rows.

Tagging.For companies we researched directly, vertical, stage, and segment come from public funding data; for the long tail surfaced via the aggregator, vertical is assigned by a keyword classifier over company and title (so expect some sitting in the catch-all “Enterprise software” bucket) and segment defaults to “Startup / scale-up.” Region is normalized from the posting location; multi-location postings count once. Seniority is inferred from the title only. Salary is captured only where the posting discloses a base range, 260 of 991 postings (26%), and is base only. Equity and bonus are almost never posted, so every pay figure here understates total compensation, severely so at the frontier labs, where equity is the majority of the package (see the note on equity above).

Caveats. This is a census of postings, not budgeted headcount. One posting can represent many seats (Palantir) or an evergreen pipeline (labs). Treat every count as a floor: broad-match searches return far more (Indeed showed ~2,150 US results and LinkedIn 1,000+ in the past month for the quoted phrase), but we include only title-matched, individually linked postings. Engineering-track titles only, so forward-deployed marketer/PM/sales/operator roles are excluded, as are FDE-equivalent roles under other names (Cognition’s and LangChain’s “Deployed Engineer,” Sierra’s “Agent Engineer”). Postings mirrored on LinkedIn or Indeed for companies already covered through their ATS were dropped as duplicates. Several VC boards (Sequoia, a16z, Lightspeed, Greylock) are client-rendered and could not be crawled; Workday-hosted sites and Google’s careers system are only partially covered. Postings expire, and links were live on the observation date. Reuse with attribution: hello@joinplank.com.

Companion analysis: The State of FDE as a Service, 2026.

Put an embedded AI team on your roadmap

Forward-deployed engineers to deploy, AI-native engineers to build, and on-demand QA pods to validate, embedded with your team, starting the same day.