What is a forward deployed engineer?
Without one, an AI product tends to stop at the demo. The vendor ships a model, and your team is left to wire it into systems the vendor has never seen.
Forward deployed engineer
Forward deployed software engineer (FDSE)FDEDelta (Palantir's internal title)
A forward deployed engineer (FDE) is a software engineer who works inside a customer's company, building and shipping their employer's product on the customer's own systems and data. FDEs stay with the account for a set period and write production code rather than advice.
The word 'forward' is borrowed from the military, where forward-deployed units sit at the front rather than at base. The engineer is the vendor's front line. They sit with your operations people, read your data as it really is, and ship working software against it.
Wikipedia's entry, read 2026-09-01, puts it this way: 'a customer-facing software engineer who develops and deploys software within a client company, often working alongside the client's employees for a defined period of time.' That last clause matters. An FDE isn't a permanent hire and isn't a one-off consultant. They're there until the system runs.
Where the role came from
Palantir created the role and its name. Inside Palantir the job is called a Delta. Gergely Orosz quoted the firm's own wording in The Pragmatic Engineer on 12 August 2025. 'Deltas (Forward Deployed Software Engineers) deploy our software platforms to customers. Deltas are part of Business Development, and their mandate is to achieve technical outcomes for our customers.'
Orosz reports that until around 2016 Palantir had more FDEs than 'normal' software engineers. Its platform was built from what those engineers learned on site, and many of them moved into core product work once Palantir Foundry launched. That loop, from the field back into the product, is the part most copies of the role leave out.
FDE, FDSE and 'FDE engineer': one role
Forward deployed software engineer (FDSE) is Palantir's full title for the job, and it's the form in its job listings and interview write-ups. FDE is the shorthand that travelled with the idea to other companies, and it's now the common term. 'FDE engineer' is a search phrasing, not a title; the E already stands for engineer. A forward deployed AI engineer is the same role attached to a model rather than a platform, covered on its own page below.
Why AI companies adopted the model
AI companies picked up the role because a working model isn't a working system. A language model answers well in a demo. Getting it to answer inside a hospital's records, an insurer's claims queue or a factory's ERP is integration work, and that work only happens where the data lives.
OpenAI set up its forward deployed engineering team at the start of 2025, according to Orosz's interview with Colin Jarvis, who leads it. Jarvis draws the line that matters: 'FDEs write code directly on customer infrastructure and use customer tooling.' A solutions architect builds a proof of concept on sample data. An FDE ships on the real thing.
Wikipedia lists Amazon Web Services, OpenAI and Anthropic among the firms hiring under the title. It notes that openings rose sharply between 2024 and 2025. In October 2025 the Wall Street Journal called FDEs AI startups' 'new (old) secret weapon.'
- EmbedSit with the operators; read the real data.
- DiagnoseWhere the product meets the process.
- BuildShip code on the customer's systems.
- Feed backField lessons go into the core product.
- Hand overRunbooks and training, or stay on.
The loop back into the product is what separates an FDE from a contractor.
How an FDE differs from a solutions engineer or a consultant
Companies use these titles loosely. Two tests cut through it: who writes the production code, and how long that person stays.
- 01Sales or solutions engineer: demos the product before the deal, builds a proof of concept (usually on sample data), then moves to the next prospect.
- 02Consultant: diagnoses and advises. Orosz puts it plainly: consultants 'make one-off recommendations, whereas FDEs generally work with customers, long-term.'
- 03Forward deployed engineer: writes the code that runs, on your systems, and stays through go-live.
01Is a forward deployed engineer the same as a solutions engineer?+
No. A solutions engineer works before the sale and builds a proof of concept, usually on sample data. A forward deployed engineer works after it and writes production code on the customer's own systems. OpenAI's Colin Jarvis describes FDEs as 'more hands-on' and working 'with more ambiguity' than solutions architects. Ask which repository the code will live in.
02Do forward deployed engineers travel?+
Often, yes. Wikipedia's entry notes that some engineers avoid the role because of the travel and the pressure to fix customer problems fast. Remote embedding has replaced some of the travel. The job is still defined by working inside the customer's world, and that often means being on site.
03How long does an FDE stay with a customer?+
For a set period, which is part of the definition. Wikipedia's example is an engineer embedded 'for a few months' to help a customer use an AI product. In practice the honest answer is 'until the system runs and someone on your side owns it.' A good engagement plans that handover from week one.
Some of the systems we have shipped
Related
- What does a forward deployed engineer actually do? →The week as it happens: data, operators, code, feedback.
- FDE vs solutions engineer vs consultant →A decision table: who writes the code and how long they stay.
- Can a small or mid-size business get a forward deployed engineer? →Yes, through an embedded senior engineer. What the engagement looks like.
- Forward deployed AI engineer: what changes when the product is a model →Evals, guardrails, data plumbing and handoffs.
- Hire forward deployed engineers →A senior engineer embedded in your operations, from a free scoping call.
- Staff augmentation →The engagement: senior engineers who join and stay.
- What is an AI agent? →The product most FDEs are deploying now.
