What is a Forward Deployed Engineer? The 2026 Guide

"Forward Deployed Engineer" is the job title of the moment — postings grew more than 800% year-over-year, OpenAI and Anthropic are hiring FDEs across three continents, and the median US base salary sits around $173K. But the title is borrowed military jargon, and it genuinely confuses people: is it consulting? Sales engineering? Support with better branding?

It's none of those. This guide explains what the role actually is, what a typical week looks like, what it pays, and how to get in.

The short definition

A Forward Deployed Engineer (FDE) is a software engineer who embeds with a customer to make their company's product succeed inside that customer's environment — by building real, production software on top of it.

The "forward deployed" part is the military metaphor: instead of building the platform back at headquarters, you're deployed to the front line, where the customer's data, systems, and problems live. The "engineer" part is literal: FDEs ship code. That single fact separates the role from every adjacent job title.

Where the role came from

Palantir invented the role in the late 2000s under the title Forward Deployed Software Engineer (FDSE). The insight was that Palantir's platforms (Gotham, Foundry) were too powerful and too general to sell as shrink-wrapped software — every customer needed engineers to mould the platform to their data and workflows. Rather than outsource that to systems integrators, Palantir hired its own engineers and embedded them with customers.

It worked. FDSEs became the engine of Palantir's growth, and the model stayed niche for a decade — until large language models created the same problem for a new generation of companies. A frontier model, like Foundry, is a powerful general platform that doesn't deliver value until someone integrates it with a company's actual systems. So OpenAI, Anthropic, Databricks, Salesforce, and dozens of others adopted Palantir's answer, usually under the modern title Forward Deployed Engineer.

What FDEs actually do all week

The mix varies by company, but a representative week includes:

• Discovery with customer stakeholders. Sitting with the people who own the business problem — ops leads, analysts, compliance officers — and translating what they say into something buildable.
• Data integration. The unglamorous heart of the job. Customer data lives in ancient warehouses, spreadsheets, and third-party SaaS. FDEs build the pipelines that connect it to the platform.
• Building the application. Writing the actual product on top of the platform: a claims-triage workflow on Foundry, a contract-analysis agent on Claude, a fine-tuned support pipeline on GPT. This is production code with real users.
• Shipping and iterating on-site or on-call. Demos to executives, fixing what breaks, expanding to the next team. Deployment velocity is the metric.
• Feeding the platform team. FDEs are the product organisation's best sensor network. The gaps you hit become the platform roadmap.

What the job is not: it's not writing PowerPoints (consultants), it's not building demo environments that get thrown away (sales engineers), and it's not answering tickets (support). The code ships and stays.

The skills that matter

FDE interviews screen for an unusual combination:

1. Strong general-purpose engineering. Python and TypeScript dominate. You'll touch data pipelines, APIs, front-ends, and infrastructure in the same month, so range beats depth in any single stack.
2. Problem decomposition under ambiguity. Palantir's famous "decomposition interview" exists because the core FDE skill is turning a messy, underspecified business problem into a technical plan. Every FDE employer tests some version of this.
3. Customer-facing communication. You'll explain trade-offs to executives who don't code and win trust from customer engineers who fear you're there to replace them.
4. Velocity and pragmatism. FDE work rewards engineers who ship a working v1 during the pilot, not the ones who design the perfect system for month six.
5. (For AI-native roles) LLM literacy. Prompting, evals, RAG architectures, agent frameworks — the 2026 FDE toolkit at OpenAI, Anthropic, and their peers.

What it pays

The median US FDE base salary is around $173K, with total compensation from roughly $130K (entry, consultancy) to $550K+ (senior, OpenAI). The full breakdown by company, level, and city is in our Forward Deployed Engineer Salary Guide, and live advertised ranges update hourly from company career pages.

The trade-offs, honestly

The role has real downsides you should weigh:

• Travel and customer rhythm. Even hybrid FDE roles bend to the customer's calendar. Production incidents at a customer site aren't 9-to-5 events.
• You build on someone else's platform. If you want to design distributed systems from scratch, the platform team is the better home. FDE work is integration-heavy by design.
• Context switching. Two or three accounts in parallel is common. Deep-focus engineers sometimes find the interrupt rate draining.
• Career legibility. "Forward Deployed Engineer" still confuses some hiring managers outside the ecosystem — though this is fading fast as the title spreads.

The upside: no engineering role is closer to revenue, and few are growing faster. FDE alumni disproportionately end up as founders, product leaders, and GTM executives because the job teaches you how software actually creates value inside organisations.

How to become an FDE

• From software engineering: You already clear the technical bar. Build evidence of customer-facing work — internal platform support, solutions work, a consulting stint — and practise decomposition-style interviews.
• From solutions/sales engineering: You have the customer skills; the gap is production coding. Ship real projects (open source, side products) and target companies whose SE and FDE ladders are adjacent. Our FDE vs Solutions Engineer comparison maps this transition in detail.
• From new grad: Palantir hires FDSEs straight from university and runs internships; consultancies like Deloitte staff Palantir FDE roles that serve as a training ground. AI-native companies mostly want 3+ years of experience.

Who's hiring right now

Palantir and OpenAI each list dozens of FDE roles at any given time, with Anthropic, Databricks, Scale AI, Salesforce, and a long tail of AI startups close behind. We track all of them — pulled directly from company career feeds, deduplicated, and updated hourly.

Browse open Forward Deployed Engineer jobs →