--- title: "AI engineer vs forward deployed engineer: 47 job posts" description: "What separates an AI engineer from a forward deployed engineer, counted in 47 job postings from 2026, and which one a company needs for which problem." url: "https://madsalt.tech/blog/ai-engineer-vs-forward-deployed-engineer/" author: "Denys Zaichenko" language: "en" published: 2026-10-31 --- # AI engineer vs forward deployed engineer: 47 job posts TL;DR An AI engineer builds AI features into a product; a forward deployed engineer (FDE) goes to the customer and makes the product work in their business. In 47 job postings from 2026 that I read and coded, the roles split on customer contact, not on AI: all 24 FDE postings were customer-facing against 7 of 23 AI engineer postings, 15 FDE postings required travel against none, and 20 asked the FDE to feed what they learn back to the product team against 1. Every AI engineer posting named LLM or agent work; only 13 of 24 FDE postings did. I'm Denys, a senior full-stack and AI engineer in Porto. I work as a forward deployed engineer on a subscription, so weigh my view accordingly. Many comparisons for this question come from course sellers, and their salary tables rarely name a source. So I counted what employers actually write. ## What is the difference between an AI engineer and a forward deployed engineer? An **AI engineer** builds features on top of models: retrieval, agents, evaluations, prompts, sometimes fine-tuning, shipped into the company's own product. The customer is usually internal. A **forward deployed engineer** works with an external customer, often on site, and builds whatever makes the product deliver in that customer's setup: integrations, data pipelines, custom workflows, and more and more often AI features too. What they learn goes back to the product team. I described the role in more detail in [what is a forward deployed engineer](https://madsalt.tech/blog/what-is-a-forward-deployed-engineer/). ## What do the job postings say? On 24 October 2026 I collected 24 postings titled Forward Deployed Engineer and 23 titled AI Engineer or a direct variant (Applied AI Engineer, Senior AI Engineer), from employers' own career pages and their Greenhouse, Lever and Ashby boards, and marked what each one states as a duty or requirement: ![What 47 job postings ask for. % of postings: 24 FDE, 23 AI engineer. FDE: works with customers: 100; AI engineer: works with customers: 30; FDE: feeds back to product: 83; AI engineer: feeds back to product: 4; FDE: travel required: 62; AI engineer: travel required: 0; FDE: builds LLM or agent features: 54; AI engineer: builds LLM or agent features: 100.](https://madsalt.tech/charts/en/ai-engineer-vs-forward-deployed-engineer.svg) *Figure 1. 47 postings from employers' career pages, read 24 October 2026.* | Stated in the posting | FDE (24) | AI engineer (23) | | ---------------------------------------- | --------- | ---------------- | | Works directly with external customers | 24 (100%) | 7 (30%) | | Feeds field findings back to the product | 20 (83%) | 1 (4%) | | Travel required | 15 (62%) | 0 | | Data pipelines or integrations | 21 (88%) | 7 (30%) | | Owns production deployment | 22 (92%) | 20 (87%) | | Builds LLM or agent features | 13 (54%) | 23 (100%) | | Trains or fine-tunes models | 0 | 6 (26%) | | Median minimum years of experience | 4 | 3 | Three rows separate the roles cleanly, and none of them is about AI: customer contact, the feedback loop and travel. Production ownership does not separate them at all; both are hired to ship. Two postings put the difference in one line each. Cursor's FDE posting: "This is not a demo role. You own the work end to end, from the first discovery call through launch, iteration, and post-launch support." Fluency's AI engineer posting: "The ability to tell the difference between 'looks good in the demo' and 'works in production' is essential." One owns a customer's outcome, the other owns a feature's quality. ## Is a forward deployed engineer an AI job? Not by definition. 11 of the 24 FDE postings did not mention LLM or agent work at all. Most of them were deployment, integration and data roles at defence, industrial and data-platform companies, but OpenAI's deployment company and Lovable were among them too. Anthropic's, Cresta's and Salesforce's Agentforce postings did name it. The title describes where the engineer works, not what they build. The reverse also happens. Linkt advertises an "Applied AI Engineer" who is "the technical face of Linkt to the client — engineer, consultant, and product owner in one", which reads like an FDE. Aircall's FDE posting says "this is not a software engineering role". Read the duties, not the title. ## Which one does your company need? Look at where the hard part of your problem sits. | Your situation | Closer to | | ----------------------------------------------------------------------- | -------------------------------- | | You know the feature; it must be accurate, fast and cheap enough to run | AI engineer | | You need evaluations, model choice or fine-tuning | AI engineer | | Nobody has pinned down what to build; it depends on how your teams work | FDE | | The data lives in several systems that must be connected first | FDE | | You sell software and each customer's setup needs adapting | FDE | | You want both: find the problem, then build the AI part | FDE who builds AI, or both roles | On my construction document platform the deciding part was of the second kind: the request was "an app instead of spreadsheets", and the real constraint, found by sitting with the workflow, was that each company's data had to stay its own. A pure feature engineer would have built the app as asked. Hiring either one takes months; the waits are in [hiring a forward deployed engineer takes months](https://madsalt.tech/blog/hiring-a-forward-deployed-engineer-takes-months/). The other ways to get the work done, agency, freelancer or an embedded engineer, are compared in [FDE vs agency vs freelancer](https://madsalt.tech/blog/forward-deployed-engineer-vs-agency-vs-freelancer/). ## What about pay? I left salaries out on purpose. Of the comparison pages I read, the ones with salary tables mostly gave no source, and one interview-prep site [says plainly](https://fdeinterviewprep.com/blog/fde-vs-ai-engineer) that no dataset compares the two titles like for like. If you need numbers, use one source for both titles, such as levels.fyi for the same company and level. Later I counted the base ranges that 41 employers post for the FDE title: forward deployed engineer salary. ## How I counted - Postings found with six searches on 24 October 2026 and opened in full; one AI engineer posting that did not load completely was dropped. - A category counts only when the posting states it as a duty or requirement; "nice to have" does not count. - One reader coded every posting, with no second check; borderline calls (for example a role titled FDE that says it is not engineering) were kept under their title. **Limits.** 47 postings is a small, convenience sample: one posting moves a percentage by about 4 points. US tech and venture-backed AI companies are overrepresented, and 18 postings showed no publish date. Five AI engineer postings came from consultancies; without them, customer contact on the AI side drops to 3 of 18. Read the table as "in these 47 postings", not as the market. Not sure which role your project needs? Happy to talk it through. No pitch. [Get in touch](https://madsalt.tech/#contact)