What Is an FDE Job? The Forward Deployed Engineer Role Explained

By Rome Thorndike

An FDE job is a Forward Deployed Engineer job. The shorthand "FDE" is what most candidates and recruiters use day to day. The role sits at the intersection of software engineering and customer work: you write production code, but you do it sitting next to (or on a video call with) a specific customer, building the parts of the product the customer needs that the core product team hasn't built and probably never will. The role originated at Palantir in the early 2010s and has spread to roughly 50 companies, with the steepest growth at AI labs and AI-native startups since 2024.

"What is an FDE job role" and "what is FDE job" return the same answer. Both are slightly broken phrasings of the same question. The full job title is Forward Deployed Engineer, sometimes Forward Deployed Software Engineer (FDSE) at companies that want to distinguish the engineering-heavy track. Pay is comparable across both titles.

What an FDE Job Actually Involves

The job is hard to describe in a single sentence because it varies more by employer than most engineering jobs. A few patterns hold across companies:

You ship code that runs at the customer. Custom data pipelines, API glue, model evaluation harnesses, vertical UI on top of a generic platform. The output is software, not a deck.
You sit closer to one customer than the core product team does. You attend their planning meetings. You know their data shape, their compliance constraints, the names of the people who will be using the thing you ship.
You scope the work yourself. The customer says "we want X." The product team says "we can't build X for everyone." The FDE figures out what X actually means, what's tractable inside the customer's constraints, and what the product team needs to expose so the FDE can build the rest.
You feed insights back upstream. Patterns you see across customer engagements become product features. The FDE is usually the first to notice when a custom build for one account is actually a generic need.
You communicate constantly. Status updates to the customer, scope negotiations with sales and product, technical decisions in writing for stakeholders who can't read code.

A typical Wednesday for a mid-level FDE looks like two hours of customer meetings, four hours of building (Python, SQL, occasional TypeScript or Go), an hour writing a design doc or technical update for the customer, and the rest absorbing pings on Slack and email from sales, product, and the customer team. Travel ranges from zero (fully remote AI-company FDE roles) to 40-60% (Palantir, especially government and healthcare deployments).

Who Hires for FDE Jobs

FDE Pulse currently tracks active FDE postings at 50+ companies. The biggest concentrations are at:

AI labs: OpenAI, Anthropic, Cohere, Scale AI. Postings appear directly on each company's careers site and rarely make it to LinkedIn quickly.
AI-platform companies: Databricks, Baseten, Modal, Sierra. The job is usually titled Forward Deployed Engineer or Applied AI Engineer and skews toward ML infrastructure work.
Enterprise platforms: Salesforce (committed to 1,000 FDE hires for Agentforce), ServiceNow, Rippling, UiPath. Pay is more standardized and travel less unpredictable than at AI labs.
Palantir: The original FDE shop. Government, healthcare, and finance deployments. Heavy travel and the most established new-grad FDSE pipeline in the industry.
Startups: Ramp, PostHog, Watershed, Onyx, Commure, and dozens more. Equity heavy, scope wide, you'll be one of a handful of FDEs.

The live FDE job board shows what's open right now, with filters by company stage, industry, and remote status.

What an FDE Job Pays

FDE jobs pay well, with a wide spread by company and seniority. The current base midpoint across active US postings tracked by FDE Pulse is around $180,000, with the typical band running $135,000 to $215,000. Senior and staff FDE jobs at Google Cloud, OpenAI, and Salesforce push into the $300,000-plus base range. Equity adds materially on top, particularly at AI labs (OpenAI's PPUs, Anthropic equity grants) and pre-IPO startups.

FDE jobs typically pay 10 to 25 percent more than equivalent-seniority software engineer roles at the same employer because the role demands both skill sets. The full breakdown by seniority, metro, remote status, and employer is on the FDE salary page.

FDE Job vs Software Engineer Job

A regular software engineer job is structured around building features for the product, talking to product managers and designers, and writing tests. The customer is abstract: a persona in a PRD, an analytics dashboard, a usage chart.

An FDE job is structured around solving one customer's actual problem with code, talking to people inside that customer's company, and writing things the customer will sign off on. The customer is a real person whose name you know. The trade-off: less time to go deep on a single technology stack, more variance in your day-to-day, and faster feedback when the thing you built either works or doesn't.

The detailed side-by-side comparison is on FDE vs Applied AI Engineer and our voices pages.

Is an FDE Job a Good Career Move?

For a particular type of engineer, yes, and the timing is unusually good. Three factors are pulling demand for FDE jobs up at once: AI products that need per-customer integration work, Salesforce's 1,000-FDE Agentforce commitment setting a public reference price, and a post-2024 generation of AI startups discovering they can't ship without an FDE function. The role's growth rate in 2025 was the highest among engineering job categories that FDE Pulse tracks.

It's the wrong job for an engineer who wants to spend two years going deep on one codebase, who avoids customer-facing work, or who can't tolerate scope ambiguity. It's the right job for an engineer who wants accelerated career optionality (product, management, founder paths all open from FDE), broader exposure, and pay that runs 10 to 25 percent above the equivalent SWE seat.

Frequently Asked Questions

What is an FDE job?

An FDE job is a Forward Deployed Engineer job, an engineering role where you write production code for and with a specific customer. You ship custom data pipelines, API integrations, AI deployment work, or vertical software on top of your employer's platform. The role started at Palantir and is now common at OpenAI, Anthropic, Salesforce, Databricks, Ramp, and roughly 50 other companies. Pay typically runs 10 to 25 percent above equivalent software engineer roles.

What is an FDE job role exactly?

The FDE job role combines software engineering work with customer engagement work in a single seat. A typical FDE writes Python and SQL most days, attends customer meetings several times a week, scopes deployment projects, builds and ships the code the customer needs, and feeds patterns back to the core product team. The role title is usually Forward Deployed Engineer or Forward Deployed Software Engineer (FDSE).

Is an FDE job the same as a regular software engineer job?

No. A regular software engineer ships features for the product, optimizing for many customers in the abstract. An FDE ships software for one customer at a time, optimizing for that customer's actual workflow and constraints. The coding bar is the same. The work environment, success metrics, and day-to-day rhythm are different.

What does an FDE actually do every day?

A mid-level FDE day usually breaks down into two hours of customer meetings, four hours of building (Python, SQL, sometimes TypeScript or Go), an hour of written communication (design docs, customer updates, sales context), and the rest absorbing async messages from sales, product, and the customer team. The exact mix varies more by employer than for most engineering roles.

How much does an FDE job pay?

Base salary for US FDE jobs runs from roughly $135,000 to $215,000 in the typical band, with a midpoint near $180,000. Senior and staff roles at Google Cloud, OpenAI, and Salesforce push into the $300,000-plus base range, and equity at AI labs and pre-IPO startups can double total compensation. The full breakdown by employer and seniority is on the FDE salary page.

What companies hire for FDE jobs?

AI labs (OpenAI, Anthropic, Cohere, Scale AI), AI platforms (Databricks, Baseten, Modal, Sierra), enterprise platforms (Salesforce, ServiceNow, Rippling, UiPath), Palantir, and a long tail of startups (Ramp, PostHog, Watershed, Onyx, Commure). The live job board shows what's open right now.

Is an FDE job hard to get?

Harder than a comparable SWE job, mostly because the interview process tests two skill sets at once: production engineering and customer communication. The technical bar is similar to a mid-level SWE at the same firm. The communication bar is what filters most candidates out. Strong candidates have written technical artifacts (design docs, RFCs, public talks) that prove they can explain complex systems to non-technical stakeholders.

Can a new grad get an FDE job?

Possible, not easy. Palantir's FDSE program and Salesforce's FDE I track are the most established new grad FDE paths. A few startups occasionally hire FDE I or Associate FDE. Most other FDE jobs want 3 or more years of software engineering experience first. The new grad FDE guide on this site has the full list of programs.