What is a forward-deployed engineer (FDE)?
Updated
Short answer
A forward-deployed engineer (FDE) is a software engineer who works directly with a customer to find the real problem, build a working solution inside the customer's own systems, and stay until the solution shows a measured result. The role is part engineer and part consultant. Palantir made the role well known, and many AI companies now hire FDEs to put their models and agents into production at customer sites.
What an FDE does
An FDE engagement usually follows the same loop:
- Discovery. Interview the sponsor and the people who do the work. Find the problem behind the request.
- Data. Read the logs, tickets, calls, and records. Check what people tell you against what the data shows.
- Process map. Draw what really happens, not what the procedure document says.
- Baseline. Measure the current state before you change anything.
- Build. Write production code, integrations, automations, or agents inside the systems the customer already uses.
- Measure and report. Compare the result to the baseline. Report it to the person who pays for the work.
- Feed the product. Tell the product team which needs repeat across customers.
Skills an FDE needs
Technical skills
- One general language, usually Python, plus SQL.
- APIs, webhooks, and data pipelines between systems that were not designed to talk to each other.
- Cloud deployment, logs, and basic security (access control, secrets, audit trails).
- For AI work: LLM prompts, retrieval, agents, and evaluations that show an agent is safe to ship.
Customer skills
- Discovery. Ask specific questions that make people tell you what they know. See discovery interview questions.
- Decomposition. Break a vague goal into parts you can build and test. See decomposition practice.
- Judgment about what not to build. Some steps should be deleted, not automated. See the four buckets.
- Clear writing. A short readout with numbers wins more trust than a long deck.
Why the job is hard
- The stated problem is often not the real problem. A sponsor asks for more staff or a chatbot. The cause is somewhere else.
- People hold facts back. Staff tell you the official version first. Talk about replacing people and they stop talking.
- The systems are messy. You build inside tools the customer already pays for, with their limits.
- You must prove the result. Without a baseline, you cannot show that your change helped.
These skills are hard to learn from reading. You learn them from engagements, and the first engagements are where the expensive mistakes happen.
Where FDEs work
- Platform and data companies that deploy their product into large customers.
- AI companies that put models and agents into customer workflows. Some use other titles, such as applied AI engineer.
- Independent practices that do FDE-style work for small and mid-sized businesses: automation, voice agents, and workflow fixes.
Questions
Is a forward-deployed engineer a sales role?
No. An FDE is an engineer. An FDE sometimes helps before a sale, but the main work is after the sale: building and deploying production software at the customer and proving that it works.
Do forward-deployed engineers write production code?
Yes. That is the main difference from a sales engineer, whose demos are usually not meant for production. An FDE's code runs in the customer's environment.
What does FDSE mean?
FDSE means forward-deployed software engineer. It is Palantir's title for the role. FDE and FDSE describe the same kind of work.
Can I learn the FDE role without a job?
You can learn the technical skills on your own. The customer skills need practice with people who do not volunteer the facts. A simulator such as FDE Workbench gives you a full engagement with interviewable characters, real-looking data, and a score.