Vibe Engines Learn
FDE vs other roles  /  vs AI Engineer
Career Track~11 min readUpdated Jul 2026
Comparison 02

FDE vs AI
Engineer

These two open the same tabs. Retrieval, agents, evals, context windows, latency budgets. The difference is what stands between them and success: for the AI engineer it is the system; for the FDE it is the customer’s organisation. And the titles are actively merging.

01

Where they overlap — which is most of the toolbox

Start with the honest part: the technical surface is nearly identical. Both roles ship retrieval over messy documents, agents that drive a multi-step workflow, tool schemas, guardrails, latency and cost budgets, and evals that decide whether any of it is good enough. Anthropic’s forward-deployed job description lists MCP servers, sub-agents and agent skills as deliverables10 — a list that would sit unchanged in an AI engineer JD.

RAG & chunkingAgent loopsTool schemasEvals & judgesPrompt/context engineeringGuardrailsLatency & costObservability

So anyone telling you FDEs are “less technical AI people” is describing a pre-sales role, not this one. The split is not depth of AI knowledge. It is what the AI is embedded in.

The distinction that actually holds

An AI engineer owns a system that many users touch: make retrieval better, make the agent more reliable, make it cheaper, ship it behind a flag. The hard problem is technical, and progress is measurable in an offline eval suite.

A forward deployed engineer owns an outcome inside one customer: get access to the data, agree what “correct” means with the people who will be blamed if it is wrong, deploy where their security team allows, and get the workflow to change. The hard problem is usually organisational, and progress is measured in adoption.

02

Side by side

DimensionAI engineerForward deployed engineer
OwnsAn AI system / featureA customer outcome that happens to use AI
DataYours; you can reshape the pipelineTheirs; undocumented, inconsistent, access-controlled
Eval setYou curate it from product trafficYou have to negotiate it with domain experts who disagree
Biggest blockerModel quality, cost, latencyData access, security review, a sceptical stakeholder
Iteration loopOffline eval → A/B → shipDemo → reaction → rescope → demo
Failure looks likeMetric plateausGreat metrics, nobody uses it
TravelRareCommon; some roles contractually 25–50%
Deploy targetYour cloudTheir VPC, on-prem, or an air-gapped enclave29
Depth vs breadthDeeper on models and inferenceBroader: infra, data plumbing, security, commercial context
→ The tell in an interview

Ask both roles to describe their hardest recent problem. An AI engineer tells you about a retrieval failure mode or a cost cliff. An FDE tells you about six weeks of not being allowed to touch the data, and what they built in the meantime to keep momentum.

03

Where the titles are merging: Applied AI Engineer

The market is converging on hybrid titles faster than the job boards can keep up. This is where the early-mover advantage is.

Watch the names. Applied AI Engineer, Agent Engineer, Forward Deployed AI Engineer and AI Solutions Engineer all describe the same emerging job: an AI engineer who is pointed at a specific customer. Anthropic’s forward-deployed role sits under Applied AI.10 Sierra hires “Agent Engineers” whose day is building and tuning customer-specific agents.28

Applied AI Engineer

AI engineer, customer-pointed

Usually a builder FDE with an AI-heavy mandate at a lab or model provider. Interview loops mix real coding with case and empathy rounds. If the AI work is what draws you and the customer work is acceptable, this is the sweet spot.

Agent Engineer

The narrower, newer one

Builds and tunes agents against a customer’s workflows — tools, guardrails, escalation paths, failure handling — and owns their behaviour in production.28 Interviews lean heavily on incremental coding and debugging planted failures rather than algorithms.

Deployment Strategist

The other direction

Palantir’s non-engineering half of forward-deployed work: owns problem framing and stakeholders, may write little production code.13 If you are an AI engineer, this is not the adjacent move — it trades code for organisational leverage.

The strategic read: the merged titles are where hiring is growing fastest and where candidate supply is thinnest, because they demand a combination — frontier-AI fluency plus enterprise delivery nerve — that few people have deliberately built. Postings for the family grew 729% year over year on one index.1

04

The skill an FDE has that most AI engineers do not

If you take one thing from this page: negotiated evals.

An AI engineer usually gets to define correctness. You pick the benchmark, you curate the golden set, you decide the threshold. An FDE almost never does. The customer’s domain experts define correct, and they disagree with each other — two senior claims adjusters will label the same case differently, and both will be adamant. Turning that into a gate a contract can rest on is the highest-leverage thing an FDE does, and it is the most-cited hard skill across frontier-lab job descriptions.

What “negotiated evals” looks like in practice
Sit two experts in a room with 40 real cases. Have them label independently. Measure their agreement with each other first — if humans agree only 70% of the time, no model is getting to 95%, and now you have a defensible target. Resolve the disagreements into a written rubric. That rubric becomes the LLM-judge prompt, the golden set becomes the regression suite, and the CI gate becomes the acceptance criterion in the statement of work. The artefact is as much a commercial document as a technical one.
Drill it

From “it should be accurate” to a CI gate

The eval track on this site walks the full path: pick metrics for a fuzzy goal, label a golden set, calibrate a judge against human grades, then wire the gate.

Evals for Client Work Eval Builder Lab Agent Evals Evals in CI
05

Comp, ceiling and exits

AI engineerForward deployed engineer
Comp levelHigh; tight-ish distribution at a given companyComparable to higher, much wider spread318
Ceiling pathStaff / principal on technical depthStaff FDE, then deployment leadership, product, or founding engineer
Best exitsResearch-adjacent, infra, ML platformProduct management, solutions leadership, founding — you have seen 20 customers’ problems up close
RiskCommoditisation of the model layerCoding atrophy; drift into services work8

The FDE’s structural advantage is pattern volume. Sit inside twenty enterprises and you accumulate a map of which problems recur, which are worth productising, and what customers will actually pay for. That is founder fuel, and it is why the exit list skews entrepreneurial. The AI engineer’s structural advantage is compounding depth in a field where depth is scarce and expensive.

Which should you pick?
Pick AI engineering if your best days end with a metric moving and you resent context-switching. Pick FDE if your best days end with a person’s work getting easier and you are bored by a quarter spent on one system. Both are AI careers; only one of them makes you good at enterprises — and, currently, only one of them is short-staffed by an order of magnitude.
Frequently asked

Quick answers

What is the difference between a forward deployed engineer and an AI engineer?

They use nearly the same toolbox — retrieval, agents, tool schemas, guardrails, evals — but solve different hard problems. An AI engineer owns a system many users touch and optimises it technically. A forward deployed engineer owns an outcome inside one customer, where the blockers are usually data access, security review and stakeholder trust rather than model quality.

Is Applied AI Engineer the same as a forward deployed engineer?

Usually yes. “Applied AI Engineer” is the frontier-lab framing of forward-deployed work — Anthropic’s forward deployed engineering sits under Applied AI. “Agent Engineer” is the narrower variant focused on building and operating customer-specific agents, and “Forward Deployed AI Engineer” is the same job with the older label.

Can an AI engineer become a forward deployed engineer?

Yes, and the technical half transfers completely. What has to be built is the customer half: scoping an underspecified ask, negotiating what “correct” means with domain experts who disagree, deploying into a VPC or air-gapped environment you do not control, and demoing to non-engineers. Negotiated evals is the highest-leverage of these.

Do forward deployed engineers do real AI work or just integration?

Both, in builder roles. Anthropic’s forward-deployed job description lists MCP servers, sub-agents and agent skills as deliverables, and evals are typically the contractual acceptance test. Integration work is real and substantial, but describing the role as integration-only fits the pre-sales-shaped roughly 30% of postings, not the builder majority.

Receipts

Sources

Every number on this page traces to one of these. Where a figure is self-reported or crowd-sourced rather than first-party, it is labelled inline.

Cited on this page

  1. Bloomberry — “I analyzed 1,000 forward deployed engineer jobs” — posting growth, builder/pre-sales/internal split. bloomberry.com/blog/i-analyzed-1000-forward-deployed-engineer-jobs-what-i-learned/
  2. Plank — FDE job-market census — 1,206 strict-definition postings across 669 companies; median posted base $185K. www.joinplank.com/fde-job-market
  3. a16z — Services-Led Growth — the margin-for-moat thesis behind FDE hiring. a16z.com/services-led-growth/
  4. Anthropic — Forward Deployed Engineer, Applied AI (JD) — 40/30/30 split; MCP servers, sub-agents and agent skills as deliverables. jobs.menlovc.com/companies/anthropic/jobs/69674588-forward-deployed-engineer-applied-ai
  5. Palantir blog — “Dev versus Delta” — FDSE and Deployment Strategist as distinct forward-deployed roles. blog.palantir.com/
  6. Perspective — 2026 Forward Deployed Engineering compensation report (1,200 FDEs) — self-reported survey data; treat as medium confidence, not first-party. www.getperspective.ai/blog/2026-forward-deployed-engineering-compensation-report-1200-fdes
  7. Sierra — “Meet the AI agent engineer” (first-party) — sierra.ai/blog/meet-the-ai-agent-engineer
  8. Anthropic — Forward Deployed Engineer, Federal Civilian (JD) — air-gapped and IL-level deployment requirements. jobs.menlovc.com/companies/anthropic/jobs/76278352-forward-deployed-engineer-federal-civilian
FDE vs AI Engineer · part of the FDE career track · Vibe Engines · 2026
Finished this one? 0 / 197 Handbooks done

Explore the topic

See this alongside everything else on the same subject — handbooks, system designs, challenges and tools, in one place.

Interview Prep Forward Deployed Engineering

More Handbooks

Explore more from Vibe Engines

Roadmaps →
System Designs →
AI System Designs →
Algorithms →
Labs →
Paper Breakdowns →