What are required for Forward Deployed Engineers?
What 870 open Forward Deployed Engineer postings and their in-house equivalents across 330 companies actually ask for · as of July 19, 2026
Forward Deployed Engineer (FDE) is the job title of the moment in AI. Across the top AI companies and AI-adjacent job boards we track, 870 open postings carry a Forward Deployed or equivalent title, spread over 330 companies. Palantir invented the role, and it is not even the biggest poster anymore: Databricks (91 open roles) and Google (71) both have more openings than Palantir (68), with OpenAI (38) close behind. Google has gone furthest in institutionalizing it, with a leveled Forward Deployed Engineer I, II, and III ladder inside Google Cloud Consulting. No big poster is as FDE-centric as the inventor, though, with 24.5% of Palantir's own open roles forward deployed. Anthropic brands the same job Applied AI Architect, and counting that org puts it fifth.
Who is hiring Forward Deployed Engineers
The twelve biggest FDE posters as of July 19, 2026: open FDE postings, with the share of the company's own open roles in parentheses. The bottom bar is not to scale.
The FDE role is not always under one title. Plain Forward Deployed Engineer covers 590 of the postings (293 companies), and Google levels it I through III. Palantir's canon is Forward Deployed Software Engineer (72 postings, mostly its own board), Databricks leads the Forward Deployed AI Engineer variant (27), and Anthropic's Applied AI Architects and Engineers (24) are the biggest equivalent brand. The rest is org structure and sprawl: 77 postings for managers, directors, and heads of forward deployment, 15 program, delivery, and recruiting roles, and 29 non-engineering variants, from Forward Deployed Product Managers to a Forward Deployed Marketer, a Forward Deployed CTO, and a Forward Deployed Staff UX/UI Designer.
One job, many names
Open postings per title family. Each row notes the family's dominant poster, or how many companies use the name.
What do the postings have in common? 77% ask for customer-facing experience, 68% set a years-of-experience bar, and 59% want Python or JavaScript in the requirements, which makes the composite profile a customer-facing generalist who codes. The striking gap: only 24% mention LLM or generative AI experience. The signature role of the AI boom mostly does not require AI experience, it requires sitting with a customer and shipping.
What FDE postings ask for
Share of the 857 open postings with parsed requirements that mention each theme; the number in parentheses is the count of postings.
The differences are where it gets interesting. OpenAI's FDE postings are the purest expression of the customer-engineer: 97% customer-facing and 55% LLM experience. Anthropic goes further: 100% of its Applied AI postings ask for LLM or generative AI experience, the only company in the table at 100%, with no degree requirement and no travel requirement. Palantir describes a different working scenario: only 25% of its postings ask for customer-facing skills, while 79% demand coding, 53% a security clearance, and LLM experience appears in 3%. A degree requirement is the big-company tell: Google, Microsoft, and C3 AI each require one in 100% of postings, against 3% at OpenAI. Travel splits the field too: 78% of Databricks postings mention it, but 0% of Google's. And Anduril is the outlier in the other direction: 0% customer-facing, 100% clearance, 100% travel. Its Forward Deployed Engineer climbs radar towers and repairs generators in the field.
The same title, different asks
Share of each company's open FDE postings that ask for each theme. Darker cells mean the theme appears in a larger share of that company's postings.
| Company | Customer facing | Python / JS | LLM / GenAI | Travel / on-site | Clearance | Degree |
|---|---|---|---|---|---|---|
Databricks 91 roles | 92% | 70% | 13% | 78% | 0% | 13% |
Google 71 roles | 82% | 85% | 31% | 0% | 3% | 100% |
Palantir 68 roles | 25% | 79% | 3% | 43% | 53% | 16% |
OpenAI 38 roles | 97% | 50% | 55% | 13% | 3% | 3% |
Anthropic 23 roles | 91% | 61% | 100% | 0% | 9% | 0% |
Cohere 13 roles | 85% | 54% | 46% | 0% | 0% | 0% |
Microsoft 12 roles | 92% | 25% | 8% | 25% | 0% | 100% |
Scale AI 12 roles | 75% | 58% | 33% | 8% | 17% | 33% |
C3 AI 11 roles | 45% | 45% | 0% | 55% | 9% | 100% |
Cresta 10 roles | 60% | 60% | 70% | 10% | 0% | 70% |
Snowflake Inc. 6 roles | 100% | 50% | 67% | 67% | 0% | 83% |
Tribe AI 9 roles | 78% | 22% | 22% | 0% | 0% | 0% |
Anduril 2 roles | 0% | 0% | 0% | 100% | 100% | 0% |
The pay bands split the same way. Across individual-contributor engineer postings with a published US base salary, the same job spans $90K at the bottom of Palantir's band to $440K at the top of xAI's. OpenAI posts up to $385K and Anthropic up to $380K for its Applied AI Architects; even Handshake, an AI data-labeling business, posts up to $375K. At the other end sit C3 AI ($120K to $158K), the company that scaled the title right after Palantir, and Anduril ($113K to $155K). The title travels; the compensation does not.
What it pays
USD base-salary bands across individual-contributor engineer FDE postings with published pay
A note on counting: we match Forward Deployed titles plus the in-house brands that name the same function (Anthropic's Applied AI Architects and Engineers, Fireworks' AI Field Engineers, Cognition's customer-side Applied AI Engineer). Anthropic's postings describe their architects as pre-sales advisors who prototype and guide deployments, so they sit closer to the sales line than an OpenAI or Palantir FDE; they are still the closest thing Anthropic posts to the role. Generic solutions-architect and solutions-engineer titles are excluded from the analysis.
Explore the interactive infographic for the full breakdown.
Want recruiting signals like these? Timeplot tracks hiring momentum, freezes, and role shifts across thousands of companies. Get in touch.
Source: Timeplot, based on public job postings from the companies we track, spanning top AI labs, AI-adjacent software companies, and startups. Data as of July 19, 2026.