In August 2020, when Palantir filed to go public, it described some of its engineers in an unusual way. Its forward deployed engineers, the filing said, had travelled to bases in Afghanistan and factories in the industrial Midwest to get its software running.
Six years later the title shows up at OpenAI, Databricks, Figma, Salesforce and others, and in August 2026 LinkedIn ranked Forward Deployed Engineer as the third most common AI job in US job postings.
What is a forward deployed engineer?
A forward deployed engineer (FDE) is a software engineer who works directly with a customer, often inside the customer's team and systems, to make a product solve that customer's problem. They help scope the problem, build the code that connects the product to the customer's data, ship it to production and stay until people use it.
Palantir, which calls the job Forward Deployed Software Engineer (FDSE), puts the difference neatly. A software engineer builds one capability that many customers use. A forward deployed engineer's focus is "one customer, many capabilities" (Palantir blog, 2019). Each FDSE works in a team that supports one customer and measures success by its impact on that customer's goal.
Where did the forward deployed engineer role come from?
A current Palantir FDSE job ad presents the role as the original version of the job and says the company pioneered embedding engineers with customers. It gives no year. The earliest dated clue we found is a 2019 Palantir blog post in which one FDSE says the job title had not changed in seven years at the company, so it was in use by about 2012.
Inside Palantir these engineers are called Deltas, a name from the early days, when each business development team was named after a letter of the NATO alphabet. Its software engineers are called Devs. In 2019 these were its two biggest engineering roles by headcount. Deltas sat in Business Development, Devs in Product Development.
Deltas also shape the product: the 2019 post says they often contribute code back to the core product. In its 2020 filing, Palantir told investors that whether customers buy more depends partly on FDEs helping them find new uses.
Why are AI companies hiring forward deployed engineers now?
AI products often pay off only once someone connects them to one customer's systems: a bank, for example, still has to wire a model into its own data, test it on its own cases and get its staff to use it. LinkedIn describes the Forward Deployed Engineer as a role that helps organisations put AI solutions in place. The hiring data points the same way:
- LinkedIn says US job postings for AI roles roughly doubled between 2023/2024 and 2025/2026, and that companies now hire people to deploy and apply AI, not only to build it.
- Its AI Talent Divide report found that, in US postings that list pay, pay rose fastest for two AI jobs: Member of Technical Staff and Forward Deployed Engineer.
- SignalFire, a venture capital firm that tracks people's roles rather than job postings, found that Forward Deployed Engineers' share of engineering roles rose 30% since 2022, against about 11% for sales engineers.
- In our count of Hacker News "Who is hiring?" posts, the share that mention "forward deployed" went from none in 2023 to 3.55% in 2026 (January to October).
- Anywhere in the post
- In the first line, where most posts list their roles
Mentions are not all openings, but in a random sample of 30 of the 2026 posts, 26 were hiring for a forward deployed engineering role.
At some companies the role is a large share of hiring. On 29 Oct 2026, our own count of public company job boards found "Forward Deployed" (or "FDE") in the title of 75 of Palantir's 304 open job ads, 114 of Databricks' 902 and 32 of OpenAI's 811 (each city counts as one ad; managers and strategists included). At Ramp, which builds software for finance teams, the forward deployed team took shape in fall 2023; by August 2025 it had grown from 2 to 16 engineers in about a year and a half. More in tech hiring trends in 2026.
What does a forward deployed engineer do all day?
The job ads give no timetable, so here is a composite day built from job ads and company write-ups.
- Morning: scoping. Agreeing with the customer which problem matters and what data exists. Palantir's FDSEs join the first customer talks to scope the problem. Ramp's motto is "always be scoping".
- Midday: building inside their systems. The pipeline, integration or AI agent that ties the product to the customer's data. Palantir and Salesforce pair FDEs with Deployment Strategists; at Palantir the strategist finds the right data in the customer's problems and runs training for users, while the engineers build the data pipeline.
- Afternoon: going live. One Palantir FDSE described being with users when the software goes live, then helping the client's IT team build on it alone.
- End of day: feeding the product. Writing up what broke and what customers keep asking for. Figma asks its FDE to turn what they learn into a reusable path for future customers, instead of permanently owning the customer's code. OpenAI wants feedback that changes product and model plans.
About half the FDE ads we read mention AI agents, and FDEs ship code into other people's systems, so checks on AI-written code matter even more; see what is harness engineering.
How much do forward deployed engineers travel?
Often, though it varies. In job ads live on 29 Oct 2026:
- OpenAI: 23 of its 25 differently titled FDE ads state up to, or about, 50%.
- Palantir: from up to 25% for its Chicago role, to 25–50% in many ads and 25–75% for some government roles.
- Scale AI (public sector): 25%, up to 50% when needed.
- Databricks: often 20%, though its London role says up to 50%.
What skills and experience do forward deployed engineer jobs ask for?
We read 147 Forward Deployed Engineer job ads at 18 companies on 28 Oct 2026, keeping engineering titles only, and counted how often each requirement appears.
- People and ownership
- Technical skills
- A degree
Four of the five most common asks are about people and ownership, not a language. Python is the one technical skill near the top. Only 14% of the ads mention a degree. Lower down come documentation and handover (46%), comfort with ambiguity (39%), prototypes (33%) and feeding lessons back to the product (30%).
How many years of experience do the ads ask for?
Of the 147 ads, 88 state a minimum. The median minimum is 5 years; three in four of those 88 ask for 3 to 8, and Staff titles start at around 8. Individual ads, all live on 29 Oct 2026, show the range:
| Level | Example ads | Experience asked |
|---|---|---|
| Student | Palantir FDSE internship (Paris) | Graduating in 2027 |
| Junior | Palantir FDSE (Chicago) | 1+ years after college |
| Junior | Baseten FDE (Training); Baseten FDE | 1–2 years; 2+ years |
| Mid-level | Ramp; Modal; Notion (India) | 3+ (preferred); 3+; 3 to 6 |
| Senior | OpenAI (London); Figma; Scale AI | 5+ (OpenAI: including customer-facing work) |
| Senior | Vercel | 6+, including 2+ customer-facing |
LinkedIn's Economic Graph report counts the people who actually started the job in the US, and Gen Z is more than two-thirds of its hires.
How is a forward deployed engineer different from a solutions engineer or a consultant?
Companies file the role in different places. Salesforce files its Forward Deployed Engineer ad under Sales. Figma calls its FDE a senior engineering role, "not a sales or services function". Vercel puts it in Professional Services. What the person does all day tells you more:
- Sales engineer: sells technical products and explains them to buyers, in the US Bureau of Labor Statistics definition.
- Solutions engineer or architect: works around the sale, with demos, proofs of concept and advice. At Vercel the Solutions Architect runs proofs of concept, while the FDE ships production code into customers' systems; at Ramp the AI Solutions Engineer carries a sales quota.
- Consultant: SignalFire says FDEs act like consultants. Palantir calls that a common misconception: consultants usually hand over a one-time analysis, while its FDSEs build lasting solutions and do far more engineering.
- Product engineer: builds the company's own product for its users. PostHog describes product engineers as full-stack developers who talk to users and decide what to build. See what is a product engineer.
A separate scan of every posting on 294 public company job boards (12 Oct 2026, same keyword rules for each title) shows the split. It used broader forward deployed titles, so compare its columns with each other, not with the 147-ad figures.
| Ads that mention… | Forward deployed | Solutions Engineer | Product Engineer |
|---|---|---|---|
| Working with customers | 71% | 68% | 31% |
| On site or travel | 55% | 18% | 14% |
| Pre-sales or proofs of concept | 25% | 74% | 11% |
| TypeScript or React | 32% | 5% | 66% |
| Design or UX | 15% | 11% | 51% |
| Median minimum years | 5 | 4 | 5 |
Source: our own count of public company job boards, 12 Oct 2026.
Solutions Engineer ads mention customers about as often as forward deployed ads, but far more often mention pre-sales work. Forward deployed ads lead on travel and on-site work; Product Engineer ads on TypeScript, React and design. For the full picture, including AI engineers, see product engineer vs forward deployed engineer vs AI engineer.
Who is the forward deployed engineer role a good fit for?
The role is likely to suit you if:
- you would rather solve one real problem completely than polish one part of a big product;
- you are as comfortable on a call with a customer's managers as in a code review;
- you can start without a clear spec;
- you write Python well, can work across the stack and don't mind some travel. OpenAI, for example, asks for full-stack production code in Python, JavaScript or similar.
Ramp's team hints at the profile: 7 of its 16 FDEs had started companies of their own. If you want long, quiet stretches on one codebase, or no travel, product or AI engineering may fit better. See how to become a product engineer and what is an AI engineer.
The takeaway
A forward deployed engineer's product is the customer's result. When you hire for it, look for production code, the confidence to talk to a customer's executives, and the habit of writing things down so the customer can carry on alone. And if an AI pilot works in a demo but stalls in your own systems, the missing piece may be an engineer who works inside your team.
Frequently asked questions
What does a forward deployed engineer do?
They work with a customer, often inside the customer's team, to scope a problem, connect a product to the customer's data and systems, and take it live. What they learn goes back to the product team.
Is a forward deployed engineer a sales job?
It sits close to sales, but it is an engineering job. Salesforce files its Forward Deployed Engineer ad under Sales, while Figma calls its FDE a senior engineering role inside product and engineering. Either way, the FDE writes production code.
How many years of experience do forward deployed engineer jobs ask for?
In our read of 147 job ads on 28 Oct 2026, 88 stated a minimum, and the median minimum was 5 years. Junior routes exist: Palantir posts FDSE internships, and Baseten has an FDE role asking for 1 to 2 years.
How much do forward deployed engineers travel?
It depends on the company. In ads live on 29 Oct 2026, most OpenAI FDE ads stated up to or about 50%, Databricks often said 20%, and some Palantir government roles said 25 to 75%.
Sources
- Palantir Blog: Dev versus Delta: Demystifying engineering roles at Palantir, 8 Apr 2019; Who Wants to be a Delta?, 21 Oct 2021.
- Palantir Technologies Inc. Form S-1. US SEC (EDGAR), 25 Aug 2020.
- LinkedIn: New LinkedIn Research Finds Women Account for Just 26% of AI Hires as AI Jobs Surge (press release), 18 Aug 2026; The AI Talent Divide (PDF), LinkedIn Economic Graph Research Institute, August 2026.
- SignalFire's State of Talent Report 2026. SignalFire, 22 Jun 2026.
- Forward Deployed Engineering. Ramp Builders Blog, 5 Aug 2025.
- Sales Engineers, Occupational Outlook Handbook. US Bureau of Labor Statistics, 27 Aug 2026.
- What is a product engineer? PostHog, read 29 Oct 2026.
- Job ads, live on 29 Oct 2026. Palantir: FDSE Chicago, internship, Japan Government, Deployment Strategist. OpenAI London. Databricks: London, Tokyo. Figma. Salesforce. Vercel: FDE, Solutions Architect. Ramp: FDE, AI Solutions Engineer. Notion India. Baseten: FDE, FDE (Training). Modal. Scale AI.
- Nilerobot, our own counts: 147 FDE ads at 18 companies, 28 Oct 2026; titles on the Palantir, Databricks and OpenAI boards, 29 Oct 2026; 294 public company job boards, 12 Oct 2026; Hacker News "Who is hiring?" posts via the Hacker News search API, 29 Oct 2026.
Need an engineer inside your team?
Nilerobot's engineers join your team and ship with you, managed and mentored by us, the way they already work inside a global client's team in Europe.
See how we work with you →