How to use this page

  • Hiring teams — pick the role you’re hiring for, see the skills that actually matter, and search the marketplace for engineers who have them.
  • Engineers — see what hiring teams expect from each role and the skills that prove you can do the work, then list yourself with those skills tagged.
  • Beyond engineering — alongside core engineering we cover product, reasoning, and safety judgement, because shipping LLMs needs all of them.

Engineering

rag-engineer

RAG Engineer

Builds retrieval-augmented systems over enterprise knowledge — ingestion, chunking, embeddings, retrieval, re-ranking, and grounded generation with citations.

Vector searchRe-rankingChunking strategyEmbeddingsCitationsHallucination control
agent-engineer

Agent / Tool-Use Engineer

Designs multi-step agent loops that pick the right tools in the right order, manage state, recover from failures, and stay within latency and cost budgets.

Tool selectionPlanningState managementError recoveryCost controlLatency budgeting
fine-tuning-engineer

LLM Fine-Tuning Engineer

Adapts open and closed models to a domain via SFT, LoRA/QLoRA, DPO/RLHF — owns dataset curation, training runs, and eval before shipping a checkpoint to production.

SFTLoRA / QLoRADPO / RLHFData curationTraining runsCheckpoint eval
prompt-engineer

Prompt Engineer

Turns fuzzy product requirements into prompts and chains that are measurably better — owns A/B prompt experiments, structured outputs, and guardrails.

Prompt iterationStructured outputsFew-shot designA/B testingGuardrailsFailure-mode analysis
applied-ai-engineer

Applied AI Engineer

The generalist who ships LLM features end-to-end inside a product team — picks the right pattern (RAG, agent, prompt, fine-tune), wires it to the app, and owns the eval that proves it works.

Pattern selectionPromptingRAG basicsTool-use / agentsEvaluationProduction wiring
forward-deployed-engineer

Forward Deployed Engineer

Embeds with a customer or design partner to ship a working AI solution against their real data and workflow. Bridges product, sales, and engineering — speed and judgement matter as much as code.

Customer scopingRapid prototypingData wranglingIntegrationsDemo-to-prodStakeholder comms
llm-systems-engineer

LLM Systems / Inference Engineer

Owns the serving layer — throughput, latency, KV-cache, batching, quantization, and the cost-per-request that makes or breaks the unit economics.

Inference serversKV cacheContinuous batchingQuantizationGPU profilingCost-per-token

Quality

evaluation-engineer

AI Evaluation Engineer

Designs the evals that decide whether a model ships. Builds rubrics, golden sets, and LLM-judge pipelines that catch regressions before users do.

Eval designRubric writingLLM-as-judgeGolden datasetsRegression testingStatistical rigor
ai-safety-engineer

AI Safety / Red-Team Engineer

Stress-tests models and apps against jailbreaks, prompt injection, data exfiltration, and policy violations — and designs the mitigations that ship to production.

Jailbreak researchPrompt injectionPolicy designRed-team automationPII / data leakageMitigation patterns

Reasoning & Logic

reasoning-engineer

LLM Reasoning Engineer

Builds the thinking layer on top of base models — chain-of-thought, planner-executor patterns, self-critique, and verifier-guided decoding for multi-step problems.

Chain-of-thoughtPlanner / executorSelf-critiqueVerifier modelsTree-of-thoughtsDecomposition

Product & Strategy

ai-product-manager

AI Product Manager

Owns the AI roadmap end-to-end: scoping where LLMs add real value, sizing risk, picking eval bars, and translating model behavior into product decisions.

Use-case scopingAI risk sizingEval bar settingRoadmap trade-offsCost / quality callsCross-functional comms
conversational-ux

Conversational UX Designer

Designs how the model talks: tone, turn-taking, repair flows, and the logic of when an agent should ask, defer, refuse, or hand off to a human.

Tone & voiceTurn designRepair flowsRefusal logicHand-off patternsFailure messaging

Don’t see your role?

The marketplace is shaped by what hiring teams actually need. If your team is searching for a role that isn’t listed — or you’re an engineer in a specialization we haven’t broken out yet — let us know.