Skip to content

Production AI FDE job description template

A template for defining production AI work, candidate requirements and working conditions.

Copy this template into your hiring document and replace every field marked [EMPLOYER: ...]. Delete the bracketed instructions before publishing. Keep the responsibilities that describe the first project. Remove any tool or requirement the role will not use.

[EMPLOYER: company name] — Forward Deployed Engineer, Production AI

Location: [EMPLOYER: city, state or approved remote region]

Work arrangement: [EMPLOYER: office, hybrid, or remote. Include required office days.]

Travel: [EMPLOYER: expected travel range or “travel is assignment-dependent; typical range is ...”]

Employment: [EMPLOYER: full-time or other status]

Compensation: [EMPLOYER: base salary range by location, plus bonus/equity if applicable]

Work authorization: [EMPLOYER: countries or statuses supported, and whether sponsorship is available]

The role

[EMPLOYER: company name] builds [product or platform] for [customer type]. We are hiring a Forward Deployed Engineer to help [customer type] use [AI capability] in [workflow or business setting]. You will take work from [discovery/prototype/listed state] through [production milestone], working with [customer engineering, operators, product, research, security, or other teams].

This role owns [EMPLOYER: exact ownership boundary]. [EMPLOYER: state who owns commercial relationship, product roadmap, security approval, and ongoing operations when those belong elsewhere.]

What you will do

  • Lead technical discovery with [EMPLOYER: customer or internal user group] and turn the workflow into a scoped delivery plan.
  • Design and build [EMPLOYER: services, applications, agents, evaluations, or integrations] that run in [customer environment or company infrastructure].
  • Write, review, test, and operate production code in [EMPLOYER: languages and main frameworks].
  • Define how the system will be evaluated, monitored, and supported after launch, including [EMPLOYER: specific checks or handoff].
  • Work with [EMPLOYER: named teams] to resolve access, privacy, security, reliability, and rollout constraints.
  • Explain technical choices and trade-offs to [EMPLOYER: technical and non-technical stakeholders].
  • Record reusable patterns and field feedback so future deployments take less custom work.

What success looks like in the first [EMPLOYER: 90 days / six months]

  • [EMPLOYER: A scoped customer workflow or deployment plan is agreed with named owners.]
  • [EMPLOYER: A pilot or production system meets the stated reliability and evaluation checks.]
  • [EMPLOYER: Users adopt the workflow or the agreed operational result is measured.]
  • [EMPLOYER: A runbook, test plan, or reusable component is available for the next deployment.]

[EMPLOYER: Delete any outcome that cannot be observed or measured. Do not publish invented adoption or impact numbers.]

Required experience

  • Experience shipping and operating production software in [EMPLOYER: relevant environment].
  • Experience turning an unclear user or customer problem into a technical plan and a working system.
  • Experience with at least one of [EMPLOYER: Python, JavaScript/TypeScript, Java, or other actual languages].
  • Experience with [EMPLOYER: LLM or generative AI systems, evaluation, retrieval, data pipelines, or other work the first project needs].
  • Clear written and spoken communication with engineers and [EMPLOYER: customer or domain stakeholders].

Useful additional experience

  • [EMPLOYER: production LLM or agent deployment]
  • [EMPLOYER: observability, evaluation, security, or governance for AI systems]
  • [EMPLOYER: experience in the customer domain]
  • [EMPLOYER: cloud, data, or integration tools that shorten the first project]

Equivalent evidence from software engineering, implementation engineering, solutions engineering, data engineering, or another hands-on technical role is welcome when it shows the ownership described above.

How the team works

[EMPLOYER: Explain who the engineer works with each week, how deployments are staffed, how decisions are made, and what happens after production launch. State whether customer visits are occasional or routine.]

Why this role exists

[EMPLOYER: Explain the business or customer need in two or three plain sentences. Do not use “high impact” without naming the result.]

Benefits and process

[EMPLOYER: Link to benefits, accommodations, interview stages, and application instructions. State what technical evidence candidates should prepare if that information is available.]

Apply: [EMPLOYER: application link]

Related reading