Forward Deployed Engineer job description: what good postings contain

A good forward deployed engineer job description says who owns a deployment from discovery to hand-over, what the engineer builds inside the customer’s environment, how much time is spent on site, and how success is measured. The title matters far less than those four lines.

Six titles for one forward deployed engineer role

Start with the title, because it decides who finds the posting. According to Perspective AI, one role is now advertised under at least six distinct titles, including Forward Deployed AI Engineer, Applied AI Engineer, Deployment or Solutions Engineer and Forward Deployed Software Engineer. The same analysis estimates that candidates who search only the literal phrase miss roughly a third of the live market.

Employers add to the confusion. Perspective AI reports that some are relabelling solutions architect requisitions as FDE requisitions to compete for the same people, which muddies comparisons of scope and pay.

Demand figures need the same care. KDnuggets notes that growth estimates run from 350 to 1,165 per cent depending on who counts and when, and that the Indeed figure is an index against a small base on a single job board. Fortune, citing Lightcast data, reports postings up more than 1000 per cent between January and August 2026 on the same period a year earlier.

So the label is unstable. A forward deployed engineer is, in the words of an Amazon Web Services posting, an engineer who embeds inside a strategic enterprise customer to design, build, deploy and run AI-powered production systems alongside the customer’s engineers.

What is a Forward Deployed Engineer? Role, lifecycle and how FDEs are measured

Ownership across the deployment comes first

The most useful line in any posting states how much of the deployment the engineer owns. OpenAI’s London posting says the engineer owns discovery, technical scoping, system design, build and production rollout, working directly with customer engineering and domain teams. That is a full-cycle role, and the wording makes it plain.

The AWS posting for a mid-level engineer goes further into operations. Amazon’s listing asks the engineer to troubleshoot production incidents and to implement monitoring, evaluations, guardrails and CI/CD that keep AI systems healthy. A candidate reading that knows the job continues after go-live.

AIU teaches the same arc as a seven-stage deployment lifecycle.

AIU teaches the deployment lifecycle in seven stages: post-sale handoff, discovery, scoping, build, deploy, iterate and hand-back.

A posting does not need to name all seven stages. It does need to show which ends are in scope. A role that stops at build attracts different people from one that includes iteration and hand-back to the customer team, and the difference should be visible in the first paragraph, not discovered at interview.

Boundaries, travel and working conditions

State what the role is not. The Pragmatic Engineer reports that, at OpenAI, solutions architects tend to be advisory and rarely write code on customers’ infrastructure, whereas FDEs write code directly on customer infrastructure using customer tooling. Say which side of that line the role sits on.

Say where it sits in the organisation, too. The same newsletter notes that Palantir’s Deltas sit within business development. Databricks tells candidates its FDEs are billable. Salesforce says it built its FDE team from engineering, professional services and customer success. Those facts change what a candidate expects from the job.

Then give travel as a number. The Pragmatic Engineer reports that Palantir expects around 25 per cent of FDE time on site, while healthcare AI company Commure estimates up to 50 per cent. Postings vary just as much: OpenAI’s London role requires travel of up to 50 per cent on a hybrid pattern of three office days, and Databricks lists 20 per cent for one level.

Name hard constraints early. One AWS US Government posting requires an active TS/SCI clearance with polygraph. LeadDev flags heavy travel and burnout risk as live concerns, so an honest travel figure protects both sides.

Requirements that match the work

Perspective AI’s sample of roughly 1,000 postings found core engineering skills such as Python or TypeScript, SQL, cloud and containers in more than 95 per cent, AI-specific skills such as retrieval, evaluation and agent orchestration in over 80 per cent, and customer discovery language in over 70 per cent. The last group is the fastest growing, so a posting with no mention of discovery is out of step with the market.

Ask for evidence of the work itself. OpenAI asks for experience of scoping and delivering complex systems in ambiguous environments, plus experience building or deploying LLM-powered systems. Google Cloud’s posting lists experience leading technical discovery sessions with customers and building evaluation pipelines. Both ask for things a candidate can describe from a real project.

Be wary of years-of-experience filters used as a proxy for judgement. Salesforce’s leader says in the company’s own article that curiosity matters more than credentials, and that the question is whether the person can solve a customer’s problem.

AIU built the Certified Forward Deployed Engineer (CFDE)™ as the first certification standard for the Forward Deployed Engineer role, which gives a posting a shared reference for what the role covers. The course teaches and the examination certifies, and they are separate products. A posting should treat CFDE as one piece of evidence to discuss, not a substitute for a work sample.

How the posting should say success is measured

Candidates read the measures section to learn what will be asked of them. OpenAI’s posting says success is judged through production adoption, measurable workflow impact and evaluation-driven feedback that changes product and model roadmaps. Google Cloud’s describes the engineer as a feedback loop that turns field insight into product roadmap input. Both name outcomes rather than activity.

AIU teaches three measures for the role.

AIU teaches that Forward Deployed Engineers are measured on three things: deployment velocity, customer success and product feedback quality.

Accountability belongs in the posting as well. One question to ask while drafting is the Signature Test.

Who signs the risk statement before go-live?

If the posting cannot answer that, the ownership of the deployment is unclear, and the first candidate to ask will find out.

Finally, avoid invented precision. Do not publish a velocity target or satisfaction score unless the team already tracks it. A named outcome that the engineer can influence is better than a number nobody has measured.

To see how AIU teaches the role and the credential built around it, use the programme details below.

The CFDE credential and who issues it →

Frequently asked questions