TL;DR: A forward deployed engineer writes production code inside one customer's systems; a product manager decides what the product does for every customer. The forward deployed product manager sits between them, owning one account's outcome without owning the code. At Scale AI the forward deployed PM and the core platform PM post the same base range, $240,000 to $300,000, while the company's FDE posting tops out at $225,000.
Palantir, the company that made the forward deployed engineer famous, had 81 postings with "Forward Deployed" in the title on October 5, 2026, and none titled product manager. I read the boards of five AI companies that day. Scale AI and Intercom now post a third job that splits the difference: the forward deployed product manager.
- 1OpenAI had 24 forward deployed postings open against 11 product manager postings. Anthropic had the reverse mix: 6 against 14.
- 2Intercom's forward deployed PM posting sets no minimum years of experience. Scale AI asks for 6+ years for the same title.
- 3Ten of Anthropic's 14 open product manager postings carry the same base range: $305,000 to $385,000.
- 4Scale AI's public sector forward deployed PM must hold an active Top Secret clearance.
What is the difference between a forward deployed engineer and a product manager?
A forward deployed engineer (FDE) is a software engineer who builds and ships production systems inside a specific customer's environment. A product manager (PM) decides what the product should do across its whole customer base, sets the roadmap and works with engineering to ship it.
Put as questions, the FDE answers "how do we make this work here?" The PM answers "what should we build next for all of our customers?"
The postings draw the line at code. OpenAI's FDE posting in San Francisco asks candidates to "write and review production-grade code across frontend and backend". Its product manager posting for API Agents lists no coding requirement at all. Instead, that PM must "define strategic priorities and roadmap for improving agentic infrastructure for API users".
- Own discovery, technical scoping, system design, build and production rollout
- Embed with customer teams and guide adoption
- Share field feedback with Research and Product
- Travel up to 50%
- Understand the problems agent builders face, as a group
- Set strategic priorities and the roadmap
- Turn priorities into SDKs, APIs and features with research and engineering
- Based in San Francisco, no travel band listed
Both jobs touch customers, and both feed the roadmap. The unit of work differs: one deployment for the FDE, the whole product for the PM. The duties split like this across the two roles and the hybrid between them:
Our forward deployed engineer role guide covers the engineer's day, tools and career path in full. The AI product manager guide does the same for the classic PM at an AI company.
The forward deployed product manager: the job in between
A forward deployed product manager is a PM who works from inside customer deployments, alongside FDEs, and decides which customer requests should change the product. Scale AI's enterprise forward deployed PM posting opens by saying what the job is not:
"This is not a roadmap PM, a CSM, or a solutions engineer. The FDPM owns product outcomes inside a portfolio of enterprise accounts."
Scale AI, Forward Deployed Product Manager, Enterprise, posting read October 5, 2026
Intercom says the same from the other side. Its senior forward deployed PM posting for the Fin AI agent files the role under product management, "not solutions engineering, customer success, or professional services." It also admits the title is new and "still being shaped".
The two companies agree on the core task. Scale's FDPM must tell "where the product is the constraint" apart from where "execution, integration, or change management is". Intercom's must turn field lessons into "specific, evidence-backed input" rather than a list of feature requests. Both describe the same filter, run from inside the account.
They differ on hands-on work. Intercom's FDPM will "pitch and demo Fin, scope the solution, hands-on configure it (including custom development if required)". Scale's asks for technical fluency, "not deep coding, but genuine comprehension of how systems work in production." So the FDPM's code column reads "light", not "no".
Scale AI also shows where the FDPM sits in a bigger machine. Three of its postings describe one loop:
The last step belongs to a classic PM. Scale's platform PM posting says that PM watches "what FD teams are building across the application layer" and makes "the call on what moves to core". The aim is for delivery teams to stop "rebuilding plumbing on every engagement." Our forward deployed product manager role guide covers the FDPM's tools and portfolio in depth.
Skills and the coding line
Companies hire the FDE on code. Scale AI's Frontier Agents Engineer posting, filed under Forward Deployed Engineering, wants "strong Python programming skills" and fundamentals in distributed systems and system design. It also lists Docker, Kubernetes, infrastructure as code and CI/CD. The day job is agent runtimes, integrations, tracing and eval harnesses, built inside customer systems.
They hire the PM on judgment. Scale's platform PM must sequence work "with incomplete information, without thrashing", and Anthropic's PM postings ask for years of shipped products. The FDPM needs both kinds of trust. Scale wants someone who can sit with "VP/C-level" buyers and senior engineers in the same week.
The experience bar is not what most guides assume. The PM roles ask for more years than the engineer roles, and the newest title asks for the fewest:
Intercom's posting says it outright: "No minimum years required." Scale's enterprise FDPM asks for 6+ years in product, program management or customer-facing product ownership. Both accept the same feeder jobs. Intercom names technical consulting, solutions engineering and customer success. Scale names consulting, field PM and "solutions engineering with ownership".
Pay: what the postings say
At all three companies that post US pay for both sides, the product role carries the higher top of range. The gap is widest at Scale AI and narrowest at Intercom. These are base salaries for San Francisco, before equity and bonus:
Seniority explains part of it. Scale's FDE posting at $180,000 to $225,000 asks for 4+ years, against 6+ for its forward deployed PM. Its senior FDE, at 5+ years, posts $216,000 to $270,000, still below the PM range. Intercom's FDE posting pays $180,000 to $200,000 in the Bay Area, against $199,800 to $222,000 for its senior forward deployed PM.
The classic PM posts higher still at the two biggest labs:
OpenAI's FDE range overlaps the bottom of its PM range rather than matching it. For FDE pay outside the US, our forward deployed engineer rate card tracks hourly rates by market, and the FDE cost-to-hire guide breaks down the full cost of a hire.
Demand: which role companies post more
The answer depends on whether the company sells a platform or a model. Palantir's 316 postings on its Lever board included 81 forward deployed roles and not one product manager title. Anthropic, by contrast, had more than twice as many PM postings as forward deployed ones.
Scale AI posts the most forward deployed PM roles of the five. Its four forward deployed PM postings include a Head of Product Management, Forward Deployed and Strategy in London, which asks for 10+ years of product leadership, 4+ of them managing PMs. It will "decide which bespoke solutions should move from custom to core product".
The model is moving past engineering. Fortune reported in September that employers from Nvidia to Scale AI have taken it into other job categories, "including in product management and tech architecture". Our post on deployment strategists vs forward deployed engineers covers the other half of that spread, the non-engineering deployment lead.
Moving between the roles
The most common move runs from engineer to PM. Our forward deployed PM guide lists the FDE as one of three feeder roles, with associate PM and solutions engineer. The postings back that up: both FDPM postings prefer a prior embedded or customer-facing technical role.
The skills gap is specific. An FDE moving across needs prioritization and the habit of saying no to a paying customer. Scale wants an FDPM who can say no to a customer "and explain why." A PM moving the other way needs tolerance for field ambiguity, and travel. Scale's public sector FDPM must travel "to customer sites, including classified facilities".
Above both sits a familiar ceiling. The FDE ladder runs to principal FDE, solutions architect and VP of solutions engineering. Our FDPM guide puts that ladder through group PM to head of product or field CTO. Scale's head of forward deployed product role shows the two ladders can meet.
When you need an FDE, a PM, or a forward deployed PM
Hire the FDE first when the product cannot yet do what an early customer needs. Hire a classic PM when many customers use the product in similar ways. Add a forward deployed PM when several large accounts each pull the product in a different direction, and someone has to decide which pulls are real.
The top-right box is where Scale AI and Intercom sit. Both pair forward deployed PMs with FDEs on their biggest accounts. Scale's posting asks its FDPM to work with the FDEs "so the account runs as a coordinated team, not a set of parallel workstreams."
With one or two pilot customers, the bottleneck is code the product does not have yet, and that is FDE work. For LLM products, the AI forward deployed engineer is the variant OpenAI and Scale AI describe in their postings.
Hire forward deployed engineers from Asia
We match you with vetted forward deployed engineers in 24 hours, at 50-70% below US cost, with a 90-day, one-time replacement guarantee. We employ them through our own EOR in 9 Asian markets, including Vietnam and the Philippines.
Our Monthly plan is $4,999 a month for a senior engineer, all-inclusive. Tell us what you are deploying and we will send a shortlist.
Frequently Asked Questions
Does a forward deployed product manager write code?
Some do, but neither company hires them for code. Intercom's posting allows "custom development if required", while Scale's asks for technical fluency, "not deep coding". The FDE owns production code in both companies.
Can a forward deployed engineer become a product manager?
Yes, and the forward deployed PM title is the usual bridge. Scale and Intercom both list prior embedded or solutions roles as a plus. The skill to build is prioritization: deciding what not to build for a paying customer.
Are these roles remote?
Few of them are. OpenAI's FDE posting asks for up to 50% travel and three office days a week. Anthropic lists its public sector PM as remote-friendly, with travel required. Scale's public sector FDPM travels to classified sites.
