The Forward Deployed Engineer Boom: 224 Open Roles Across 39 AI Companies

For most of the last decade, the smartest engineers in tech wanted to work on the model. The frontier was the model. The status was in the model. The compensation followed the model.

In 2026, that frontier moved. The bottleneck is no longer model capability — it's getting models from a demo to a production system that meaningfully changes a customer's revenue or cost line. An MIT NANDA study published in late 2025 examined 300 public enterprise AI projects and found that 95% produced little or no measurable impact on profit and loss. The models worked. The deployments didn't.

The role that fixes this has a strange name: Forward Deployed Engineer. We track it on JobsByCulture and across our 118 culture-profiled companies. As of today, May 30, 2026, JBC's live job index shows 224 open Forward Deployed Engineer roles across 39 distinct companies. The role didn't exist on most of these career pages 18 months ago. It is now, by some external estimates, the fastest-growing AI job title in the United States.

Here's what's happening, who is hiring, what they pay, and how the role is structurally different from the engineering job you might already have.

What a Forward Deployed Engineer Actually Does

A Forward Deployed Engineer (FDE) is a technical specialist embedded inside a customer's company — either physically on-site or in tight functional partnership — to scope, build, and deploy an AI system from initial requirements through measurable business outcomes.

The role blends three job functions that used to live in three separate orgs:

• Software engineer. Production-grade Python and TypeScript, comfort with RAG pipelines, eval frameworks, agent development, and production observability. Most FDEs ship real code into customer environments.
• Solutions architect. Scoping a customer's actual problem — usually messier than the sales pitch suggested — into a system that can be built. Deciding what to automate, what to leave manual, what data to use, what tradeoffs to accept.
• Customer success. Staying on the account until the deployment hits a measurable business outcome: renewal, expansion, revenue lift, cost reduction. The FDE doesn't "throw it over the wall" to a CS manager — they own the outcome.

Read any of the open job descriptions on our engineering jobs page and you'll see the same pattern: more end-to-end ownership than a typical platform engineer, more technical depth than a typical solutions engineer, more outcome accountability than either. The combination is rare, and the compensation reflects that.

Where the 224 Open Roles Actually Are

We pulled the live data from the JBC platform on May 30, 2026. Here's the breakdown of the 224 active Forward Deployed Engineer roles by company:

Two patterns stand out. First, this isn't a one- or two-company trend — it's 39 companies across infrastructure, applications, voice, video, code, sovereign AI, and vertical-specific platforms. Second, Palantir alone has more open FDE roles than the next two companies combined — a reminder that the role has been Palantir's structural advantage for over a decade before the broader industry caught up.

Why 2026 Is the FDE Year

Job postings for Forward Deployed Engineers jumped by more than 800% between January and September 2025, per industry tracking. That kind of growth is rarely accidental. Three structural shifts drove it.

1. Enterprise AI pilots are failing at scale

The MIT NANDA study found that 95% of enterprise AI pilots fail to produce measurable P&L impact. Almost all of those failures trace to deployment, not models. Customers can't:

• Define the success metric clearly enough to evaluate against it
• Get clean enough data into the system to make outputs reliable
• Build the evals that would catch regressions before they hit users
• Bridge their existing software stack to the AI service
• Train internal teams to actually use the new capability

None of those problems are solved by a better model. They're solved by an engineer sitting inside the customer for three months who knows what good production AI looks like and has the authority to do it.

2. Palantir proved the financial case

Palantir invented the FDE model in the early 2010s and built the company around it. The result: from 2020 to 2025, Palantir's stock returned approximately 640%. That return is widely attributed to the FDE model's ability to land enterprise customers, expand inside them, and hold them through renewal at higher contract values.

For five years, this was Palantir's idiosyncratic competitive advantage that the rest of the industry dismissed as "expensive solutions consulting." In 2025, watching frontier-lab CEOs read Palantir's earnings, that dismissal became impossible to defend.

3. The frontier labs followed in May 2026

In May 2026 — the same month this article is being published — two of the largest AI labs in the world launched dedicated FDE business units within days of each other:

• OpenAI spun up "The Deployment Company," a dedicated FDE business with $4B+ in announced enterprise commitments.
• Anthropic announced a $1.5B joint venture with Blackstone and Goldman Sachs to embed Claude FDEs inside financial-services customers.

When the two largest labs in the world spin up dedicated, separately-incorporated FDE businesses in the same month, the role has stopped being a tactical hire and become a strategic category.

What Forward Deployed Engineers Make

Compensation reflects the hybrid skill set — engineering depth plus customer-facing communication plus product judgment plus business accountability. The market is paying for it.

For comparison, our AI Engineer Salary Guide shows mid-level AI Engineers in the $185K–$250K band, and senior at $250K–$400K. The FDE premium is roughly $50K–$100K at every level, reflecting the additional customer-facing accountability.

The Skill Profile That Actually Wins These Roles

Reading across the 224 active job descriptions, the requirements cluster into four areas. If you're trying to position yourself for an FDE role, optimize against these:

1. Production AI engineering. RAG pipelines that actually retrieve the right thing. Eval frameworks that catch regressions. Agent loops that don't hallucinate tool calls. Inference observability. Most ML engineers from the 2017–2022 generation are weak here — this is a new skill stack, and it rewards engineers who learned by shipping recent production AI systems, not engineers who learned by training models in a notebook.
2. Customer-facing communication. Articulating a tradeoff to a non-technical executive. Writing a project plan a customer's procurement team can sign. Saying no to a feature request without losing the renewal. This is the skill that filters traditional ICs out of the FDE pipeline.
3. End-to-end ownership. FDEs ship the integration, run the eval, train the customer's team, and stay on the account until the system is in production. Engineers who like clean handoffs to ops or CS will not enjoy this role.
4. Business judgment. Knowing which of three possible scopes will produce measurable revenue lift in 90 days, vs. which one is just technically interesting. This is the part that most senior engineers underestimate — and it's also the part that drives the compensation premium.

For engineers prepping interviews for FDE roles at these companies, our Palantir interview prep, OpenAI interview prep, and Databricks interview prep guides cover the technical and behavioral bar at the largest FDE employers.

Who Should — And Shouldn't — Pursue an FDE Path

This role is built for you if…

• You're a strong engineer who genuinely enjoys customer-facing work and finds shipping into a real account more satisfying than shipping internal infra.
• You're tired of building demos that don't get deployed, and you want to own the outcome instead of the prototype.
• You want to be in the room when a Fortune 500 decides what its AI roadmap actually is, not on a Slack thread reading about it.
• You're OK with travel — many FDE roles require regular on-site time at customer offices.

Skip the FDE path if…

• You want to maximize time on deep technical problems and minimize meetings. FDE is roughly 50% engineering and 50% customer-facing — if that split sounds painful, look at engineering-driven IC roles instead.
• You're optimizing for paper-track research output. Frontier labs (Anthropic, DeepMind) still have classical research tracks where FDEs are not the path.
• You don't enjoy ambiguity in scoping. FDE work is, by definition, undefined when you walk in. Engineers who need a clear spec will struggle.

Where the Trend Goes Next

Three predictions for the next 12 months that are worth tracking:

FDE total comp at the top of the band will hit $750K-$1M. The current ceiling of ~$600K reflects market lag. As demand keeps growing and supply stays scarce, principal-level FDEs at top labs will price into the seven-figure range to compete with the AI research track.

FDE-specialist staffing firms will emerge as a category. Palantir grew an army of FDEs in-house over a decade. New AI companies don't have time to do that. Expect a wave of FDE-specialist staffing firms, similar to the early Big-4 consulting equivalents, to fill the gap. Some of them will themselves be reasonable startups to join.

The role will fragment. Today, "Forward Deployed Engineer" is one title with wide variance. By mid-2027, expect cleaner subspecialty distinctions: FDE-Infrastructure, FDE-Eval, FDE-Agent, FDE-Sovereign. Job titles inside Palantir, OpenAI, and Mistral are already moving this direction (see the job listings on our jobs board for examples).