METAHEURISTICBook a call
[Hiring guide]

Agentic AI engineer: job description and hiring guide.

A job description you can paste and edit, the skills that actually predict success, the interview loop that surfaces them, and the five mistakes that make this hire go wrong.

Most agentic AI engineer job descriptions are a list of framework names. That screens for people who have read the documentation and against people who have been paged at 3am by a loop that would not stop. Here is a version that does the opposite.

The template

Agentic AI Engineer

You will design and ship systems where a language model plans, calls tools, and takes real actions against real data - and you will make those systems measurable, safe and cheap enough to run at our volume.

What you will do

  • Design agent control flow: state, transitions, budgets, retries, escalation to a human.
  • Build and harden tool integrations against our internal systems, with validated, idempotent contracts.
  • Own the evaluation harness: golden sets from real traffic, regression gates in CI, quality reported per release.
  • Model and manage cost and latency per completed task, not per token.
  • Build guardrails against prompt injection, tool abuse and data exfiltration, and test them adversarially.
  • Instrument everything: traces you can replay, failure taxonomies, dashboards our on-call can read.
  • Work with product and operations to decide what the agent must never decide alone.

What we look for

  • You have shipped a non-deterministic system to production and kept it there.
  • You reason about a language model as an unreliable component in a distributed system, not as an oracle.
  • You measure before you tune, and you can show us the eval set you built.
  • You are comfortable deleting an abstraction that is not paying for itself.
  • Strong general engineering: Python or Go, queues, APIs, databases, observability.

Nice to have

  • Retrieval systems at scale, including permission-aware retrieval.
  • Open-weight model deployment, self-hosted inference, or on-premise constraints.
  • Security background, or experience red-teaming model-driven systems.
  • Public writing or open source in this space.

Cut freely. Every line you cannot honestly offer costs you candidates.

The skills that actually predict success

SignalWhy it predictsHow to test it
Has built an eval setEverything downstream depends on being able to measureAsk to see one. Ask what they would label and who would label it
Thinks in budgetsLoops and cost are the two failure modes that reach the CFOAsk how they stop an agent that will not terminate
Designs tool contractsMost “the model is dumb” bugs are interface bugsGive them a bad tool signature and ask what is wrong with it
Has a threat modelPrompt injection is a design concern, not a filterAsk how a retrieved document could take over the agent
Deletes thingsAgent codebases rot into layers of unused abstractionAsk what they removed from their last system and why
Reads tracesDiagnosis beats intuition after week twoHand them a real trace with a failure in it

An interview loop that works

  1. Portfolio walkthrough, 45 minutes. A system they shipped: architecture, what broke, what they measured, what they would do differently. Push until you hit something specific. People who have only demoed cannot get past the second follow-up.
  2. Design exercise, 60 minutes. “An agent that processes incoming invoices against our policy and pays the ones under $500.” Watch for: where they put the human, how they scope tool permissions, what they refuse to automate, what budget they set, and how they would know it works.
  3. Debugging exercise, 60 minutes. Give them a real trace where an agent looped, or retrieved the wrong document, or spent forty dollars on one task. Diagnosis quality is the single most transferable signal in the loop.
  4. Code review, 45 minutes. A tool implementation with an ambiguous contract, an unbounded retry and a missing idempotency key. Do they find all three, and do they explain why each matters?
  5. Values and collaboration, 30 minutes. This role says “no” to product a lot. Make sure they can do it without being insufferable about it.

Skip the algorithm screen. It tests nothing this job requires.

Five ways this hire goes wrong

  1. Hiring a researcher for a production role. Impressive background, no instinct for on-call. Ask what they have operated, not what they have read.
  2. Screening on framework names. The frameworks in your JD will be replaced within eighteen months. The judgement will not.
  3. No counterpart on the product side. Agentic systems make policy decisions - what to automate, what to escalate. Without a product owner, the engineer ends up guessing at your risk tolerance.
  4. Hiring before the use case is proven. You will spend a quarter recruiting for a system whose requirements change in month two. Contract first if you are still learning.
  5. No budget for evals. If the first roadmap has no line item for an evaluation harness, the hire will spend a year tuning prompts by feel and you will never know whether it is getting better.

Contracting an agentic AI engineer gets a system into production while the search runs - and leaves your eventual hire with a working codebase, evals and runbooks instead of a blank repository. Twenty-minute call to talk it through.

Questions before you hire.

Is an agentic AI engineer the same as a machine learning engineer? +

No, and conflating them is the most common hiring mistake. A machine learning engineer trains and serves models. An agentic AI engineer composes existing models into reliable systems - control flow, tool contracts, retrieval, evals, guardrails, cost. The overlap is real but the day job is closer to distributed systems engineering than to modelling.

Do we need a PhD? +

Almost never. Frontier research is not the bottleneck for shipping agents; production engineering judgement is. Optimise for someone who has run something unreliable in production and made it reliable.

What should we pay? +

Benchmark against your senior backend and platform band rather than against data science, and expect a premium for people who have genuinely shipped agents rather than demoed them. Vet the difference with the portfolio question in the loop below.

Should we hire, or contract first? +

If the use case is unproven, contract first and hire once you know what the system needs to be. If it is proven and permanent, hire. Both paths are laid out on the rates and engagement models page - including the case for contracting nobody at all.

Can we just use this JD as-is? +

Yes. Copy it, cut anything you cannot honestly offer, and delete any requirement you would not actually reject a strong candidate over. Most AI job descriptions fail because they list twenty technologies and screen out the people who would have been good at the job.

Need an agentic AI engineer this quarter?

Contract, fractional, or embedded with your team. Start with a 20-minute call - you leave it with an honest read on whether your use case is buildable, whatever happens next.