As companies move from trying AI models to making them work inside real workflows, the role keeps appearing in job posts. It also keeps getting confused with consultants and sales engineers. This post explains what an FDE does, how the role differs from a software engineer, and what to look for if you need to hire one.
What a forward deployed engineer actually does
An FDE works inside the customer's world, not behind a ticket queue. The day-to-day usually follows three steps:
- Scope the problem. Sit with the customer and find what really matters, which is often different from what the first request said.
- Build and deploy. Write code that runs on the customer's own systems, against their data and constraints.
- Feed learnings back. When the core product blocks the customer, take that evidence back so the product improves.
The work is hands-on and close to the business problem. An FDE is judged by whether the customer got an outcome, not by how many features shipped.
Forward deployed engineer vs software engineer
Both roles share the same engineering base: both write, test and ship software. The difference is focus.
| Software engineer | Forward deployed engineer | |
|---|---|---|
| Customers | Many customers, one product | One customer, many needs |
| Starts from | A defined spec | A messy business problem |
| Builds | Reusable features | Tailored solutions |
| Success looks like | A shipped product | A customer outcome |
These are typical tendencies, not rules. Neither role is better, and many engineers move between the two during a career.
What an FDE is not
Two neighbouring roles are often mistaken for the FDE.
| Role | What it typically does | How the FDE differs |
|---|---|---|
| Consultant | Makes one-off recommendations | Builds the solution and stays involved long term |
| Sales or solutions engineer | Advises and demos on prototypes | Writes real code on the customer's own infrastructure |
The common thread is that an FDE is accountable for working software in the customer's environment, not for advice or a demo.
Why the role is getting attention now
AI changed the shape of the work. Companies have moved from trying AI models to making them work inside real workflows, with real data, real permissions and real users. Someone has to do that integration, and it rarely fits neatly into a product backlog.
That is where the FDE model fits. It tends to appear in organisations where integration work is heavy and customer needs differ from one account to the next, such as AI companies, consulting firms, fintech, insurance and healthcare.
Demand for the title is real, but job descriptions vary widely. When you hire, define the outcomes you need rather than relying on the label.
What it takes: more than coding skills
Great code is the entry ticket. These four things set the role apart:
- Technical depth. Builds and ships real software.
- Customer fluency. Listens, asks good questions and earns trust.
- Problem framing. Turns a vague need into a clear scope.
- Business sense. Knows which outcome matters most.
When you hire, test for the last three as seriously as the first. Give candidates a messy, realistic scenario and watch how they ask questions and narrow the scope, not just how they write code. If you only need extra capacity for a defined roadmap, a regular software engineer is the better fit. Choose an FDE when the customer's problem is the work.