What a forward deployed engineer actually does
A forward deployed engineer (FDE) is a software engineer who works directly inside a customer's world — their data, their systems, their politics — and ships the software that turns a product into a measurable outcome for that customer. You are not building a feature for millions of anonymous users. You are making one important customer successful, and feeding what you learn back into the product.
Where the role comes from
The title was popularised by Palantir, which split engineering into people who build the core platform and people who deploy it at customer sites (internally nicknamed "Devs" and "Deltas"). Many AI labs and AI-native startups have since adopted the model, because large language models are a horizontal capability: the value only appears once someone wires the model into a specific customer's data, workflows and constraints. That someone is the FDE.
What the job looks like day to day
A typical week mixes very different kinds of work:
- Discovery — interviewing users, shadowing a workflow, reading the customer's data dictionary.
- Building — writing production code: pipelines, integrations, prompts, evaluation harnesses, small UIs.
- Deploying — getting software running inside the customer's cloud, network and security rules.
- Communicating — demos to executives, status updates, unblocking access requests, writing things down.
- Feeding back — telling your own product team what customers actually need and where the platform falls short.
An FDE is measured on the customer's outcome, not on lines of code or tickets closed. "The claims team now processes 30% more claims per day" is an FDE result; "I built the claims API" is not.
How it differs from neighbouring roles
| Role | Primary customer | Writes production code? | Success looks like |
|---|---|---|---|
| Software engineer | Internal product roadmap | Yes, for the core product | Features shipped, reliability, velocity |
| Solutions / sales engineer | A prospect in a sales cycle | Demos and prototypes | Deals won |
| Solutions architect | Customer's technical team | Reference designs, some code | Customer can build on the platform |
| Management consultant | Customer's leadership | Rarely | Recommendations adopted |
| Forward deployed engineer | A specific strategic customer | Yes — in the customer's environment | Measurable outcome in production |
The FDE overlaps with all of them. You need an engineer's depth (you will debug a broken OAuth flow at 11pm), a consultant's structure (you will run a scoping workshop), and a product manager's judgement (you will decide what not to build).
The three skills that matter
- Engineering range. Not just your favourite stack: APIs, data pipelines, SQL, cloud, auth, deployment, and increasingly LLM application patterns like RAG, tool calling and evals.
- Problem framing and business sense. Turning "we want AI" into a problem with a baseline, a metric and a business case someone will pay for.
- Trust-building communication. Explaining trade-offs to a CFO, a security reviewer and a frontline user in the same afternoon — and writing it down so it survives.
If you are coming from a full-stack background, you already have more of the first skill than most candidates. The fastest way to stand out is to show evidence of the second and third: write-ups that state the problem, what you built, and the measured result.
Why companies pay a premium for it
Enterprise deals for AI and data platforms are large, and they are won or lost on whether the first deployment produces value. An FDE de-risks that. A single successful deployment can turn a pilot into a multi-year contract and produce a reference customer — so companies treat strong FDEs as revenue-critical, and compensate accordingly.
Expect "Why FDE and not a regular engineering role?" A strong answer connects to outcomes: you enjoy owning a problem end-to-end, you have done customer-facing delivery before, and you get energy from seeing software change how real people work.
Senior Full-Stack Engineer · AI & AWS. I turn ambiguous customer problems into production systems — cloud, data and AI.