What Is a Forward Deployed Engineer (FDE)?

What a forward deployed engineer does

An FDE’s job is to close the gap between what a product can do in a demo and what it does inside your business. The work usually looks like this:

• Learning the workflow. Sitting with the people who will use the product to see how work really gets done, including the exceptions nobody wrote down.
• Configuring and connecting. Setting up the product and linking it to your existing systems, often through an API (a standard way for software to exchange data), so it can read and write the data it needs.
• Building what’s missing. Writing custom code, rules, or instructions for needs the standard product doesn’t cover.
• Feeding back to the vendor. Reporting what customers keep asking for, so the vendor’s product team can build it into the product for everyone.

FDEs typically write working code and configuration, not slide decks. They are usually assigned to one customer, or a small number of customers, at a time.

Why AI companies use them

The role is most closely associated with Palantir, the data software company, which calls it a forward deployed software engineer and has long placed engineers directly with its government and commercial customers. In recent years, a number of AI companies have adopted a similar title and approach.

The reason is practical. An AI agent or custom AI system usually needs real setup work before it fits the way a particular business runs. It needs access to the right data and tools, guardrails on what it may and may not do, and instructions that reflect how that business operates. Those details differ from one customer to the next, and they are often where AI projects get stuck between a pilot and everyday use.

An FDE gives the vendor someone close enough to the customer to sort out those details, and gives the customer an expert who knows the product from the inside.

FDE vs. sales engineer, consultant, and support

These roles all put technical people in front of customers, so they’re easy to mix up. The differences come down to who employs them, when they show up, and what they produce.

• Sales engineer: works before the sale. Runs demos, answers technical questions, and may build a proof of concept to help close the deal. Usually steps back once you sign.
• Consultant: often works for an independent firm, or for themselves, and is hired by you. May work across many tools and vendors, but has no inside line to the vendor’s product team. Many vendors also have their own professional services or implementation consultants. The line between them and FDEs is blurry, but FDEs are usually engineers who write custom code and work closely with the product team.
• Customer support: responds to questions and problems after launch, usually through tickets, and generally doesn’t redesign how the product fits your operation.
• Forward deployed engineer: works for the vendor, mainly after the sale (sometimes starting with a pilot), writes and deploys working code alongside your team, and has a direct line back to the people who build the product.

Job titles vary between companies, so ask what the person will actually do rather than relying on the label.

What to ask if a vendor offers an FDE

An FDE can be a real advantage: hands-on help getting a product from signed contract to everyday use. It also shapes your costs and how much you depend on the vendor. Before you sign, ask:

• Scope: What will the FDE build, configure, or connect, and what is your team expected to do?
• Cost: Is their time included in the contract or billed separately, and is it capped?
• Duration: How long will they stay, and what support do you get after they leave?
• Handover: How will knowledge pass to your team (documentation, training, shadowing) so you can run and adjust the system yourselves?
• Ownership and access: Who owns any custom code or configuration, and which of your data and systems will the FDE be able to see?

Having an FDE is a service model, not a quality standard. It doesn’t tell you whether a product meets the security and governance expectations described under enterprise-grade AI, so check those separately. And plan from day one for the system, and the AI workflow it supports, to keep running after the FDE moves on.

Further reading

• AI ImplementationCovers SN3’s work redesigning processes, connecting systems, automating work, and deploying AI, which is useful context if you are weighing hands-on help to get AI into everyday use.
• AI EmployeesShows SN3’s role-trained AI systems deployed inside the tools a team already uses, with guardrails and support after launch, a useful comparison point when weighing how much hands-on deployment help you need.