- Signal Through the Noise

Every AI-related job title currently in use in 2026, in one reference. Use this to standardize your hiring, decode a job posting, benchmark a salary band, or figure out which titles are real work versus marketing theater.

Each role card includes: definition, day-to-day, tech stack, differentiator from adjacent roles, seniority ladder, US and EU compensation, representative companies, origin, and a verdict. Verdicts use four tags: Stable Fading Theater Predicted.

Sources: Levels.fyi Q3 2025, Glassdoor April 2026, ZipRecruiter, Indeed Hiring Lab, Cognizant press release (Aug 29, 2025), Pragmatic Engineer, Simon Willison, Andrej Karpathy, FT, LangChain State of Agent Engineering 2025, GuruSup CAIO 2026 report. All figures are total compensation unless noted.

A. Core AI Engineering Titles

Titles for engineers who build AI-powered features on top of pretrained models. The largest and most overloaded category.

1. AI Engineer #

Stable & Growing

Definition: Builds AI-powered application features, primarily using pretrained models via API. The default general-purpose title for production LLM work in 2026.
Day-to-day: Integrates LLMs, ships RAG pipelines, writes evals, manages cost and latency, owns agentic flows end-to-end.
Stack: Python or TypeScript, OpenAI/Anthropic/Gemini APIs, LangChain or LlamaIndex, vector DBs (Pinecone, Weaviate, pgvector), LangSmith/Braintrust, Docker, AWS/GCP.
Differentiator: Vs. ML Engineer, does not train models from scratch. Vs. Software Engineer, LLM-specific evals, prompting, cost engineering.
Seniority: Junior → Mid → Senior → Staff → Principal.
US comp: Levels.fyi median $159,430 for “AI Engineer”; $245,000 when tagged “AI Software Engineer.” Google L3 to L6 range $183K to $583K. Microsoft $238K to $355K (median $263K). Entry $110K to $140K base; senior $180K to $250K base.
EU comp: UK median ~£75K base; senior €95K to €160K in Berlin, Amsterdam, Paris.
Companies: Google, Microsoft, Meta, Amazon, Salesforce, Stripe, Shopify. Essentially universal.
Verdict: LinkedIn called it “the fastest-growing role in the US” in 2025. Will absorb most sibling titles over the next 18 months.
Origin: Mainstreamed by the AI Engineer World’s Fair (2023 to 2024).

2. Applied AI Engineer #

Stable

Definition: AI Engineer variant with heavier emphasis on shipping customer-facing AI into real workflows. At OpenAI and Anthropic this title often overlaps with FDE.
US comp: $200K to $500K TC at frontier labs; mid-tier $160K to $260K.
Companies: OpenAI, Anthropic, Cohere, Hugging Face, Mistral.
Verdict: Slight prestige premium over “AI Engineer” at AI labs. Same job in practice.

3. AI Software Engineer #

Stable

Definition: Software Engineer whose primary specialization is AI-powered products. Levels.fyi’s canonical bucket.
US comp: Median $245,000 TC. AI-focused engineers earn 6.2% more than non-AI at entry, widening to 18.7% more at Staff level in 2025.
Verdict: Effectively a synonym for AI Engineer with a higher comp band because the word “Software” anchors it to SWE pay.

4. Generative AI Engineer / GenAI Engineer #

Consolidating

Definition: AI Engineer scoped to generative models (text, image, audio, video).
Day-to-day: Prompt design, RAG, fine-tuning, multimodal pipelines, safety filters.
Differentiator: Vs. AI Engineer, nominally narrower, in practice identical.
Comp: $175K to $250K base; ~$220K to $350K TC.
Companies: Accenture, Deloitte, Capgemini, Infosys, TCS (enterprise services love the “GenAI” label); also Adobe, Autodesk, Canva.
Verdict: Branding variant of AI Engineer. Peaked as a distinct title in late 2024; consolidating.

5. LLM Engineer #

Real Specialization

Definition: Specialist in productionizing large language models end-to-end.
Day-to-day: RAG architectures, eval harnesses, prompt optimization, fine-tuning, context management, cost and latency tuning, guardrails.
Stack: Same as AI Engineer plus fine-tuning (LoRA, QLoRA), quantization, vLLM/TGI, deeper eval frameworks.
Differentiator: Vs. AI Engineer, more depth in the model layer. Vs. ML Engineer, no from-scratch training.
US comp: Mid $160K to $210K. Senior $210K to $320K base. Staff/Principal $300K to $450K base, TC past $1M at top tier.
Companies: Anthropic, OpenAI, Cohere, Perplexity, Character.AI, xAI, Databricks.
Verdict: Real specialization, real premium. One of the few AI titles where the extra pay is justified.
Origin: Emerged in 2023 post-GPT-4.

6. Prompt Engineer #

Declining

Definition: Designs and optimizes prompts for LLM applications.
Day-to-day: Prompt iteration, few-shot templating, chain-of-thought design, evals, red-teaming prompts.
Differentiator: Vs. AI Engineer, narrower, less code-centric. A subset of context engineering.
US comp: Glassdoor median $129,461 (range $102K to $166K). Coursera median $126,000. Levels.fyi range $83,200 to $230,000. ZipRecruiter lower at ~$63K to $87K (non-engineering variants). Senior specialists can hit $270K+.
Companies: Anthropic (famous early listing), Scale AI, Surge AI, Google, Klarna.
Verdict: Declining as a standalone title. Being absorbed into AI Engineer and Context Engineer.
Origin: 2022 (OpenAI and Anthropic early listings).

7. Context Engineer #

Inflated

Definition: Engineer who designs the full context envelope around LLM calls: retrieval, memory, tool definitions, system prompts, metadata, governance, so agents behave reliably.
Day-to-day: Designs retrieval pipelines, memory systems, tool schemas, “context packs” for reusable business logic, eval harnesses for context drift, governance of data injected into agents.
Differentiator: Vs. Prompt Engineer, prompt engineering is a subset. Vs. AI Engineer, focused on the information architecture layer, not the whole stack.
US comp: Not yet on Levels.fyi. Tracking with AI Engineer bands: $140K to $240K base at tier-1 enterprises.
Companies: Cognizant announced 1,000 Context Engineer hires on August 29, 2025 with Workfabric AI. Also Accenture, Deloitte, Infosys. Shopify CEO Tobi Lütke publicly defined context engineering; Karpathy endorsed the term in June 2025.
Verdict: Real discipline, inflated title. The work is legitimate but the role is largely consulting-industrial branding tied to a partner platform sale.
Origin: Karpathy and Lütke, mid-2025. Operationalized by Cognizant Aug 2025.

8. Agent Engineer / Agentic AI Engineer #

Growing

Definition: Builds autonomous or semi-autonomous AI agents that plan, use tools, and execute tasks.
Day-to-day: Multi-step planning loops, tool-use frameworks, agent memory, orchestration (LangGraph, CrewAI, AutoGen, OpenAI Agents SDK), error recovery, observability.
Differentiator: Vs. AI Engineer, focuses on multi-turn autonomy. Vs. LLM Engineer, more orchestration, less model internals.
US comp: Glassdoor “Agentic AI Engineer” average $191,434 (range $151K to $246K). Staff/Principal at leading agent shops: $350K to $500K TC.
Companies: Anthropic, OpenAI, Cognition Labs (Devin), Salesforce, ServiceNow, Zapier, LangChain, Sierra.ai.
Verdict: Real work, fuzzy distinction from AI Engineer. LangChain’s State of Agent Engineering 2025 showed 986% YoY growth. Likely stabilizes into “AI Engineer (Agents)” rather than its own permanent bucket.
Origin: 2023 to 2024, accelerated after Claude’s tool use and OpenAI Assistants API.

9. RAG Engineer #

Fading

Definition: Specialist in retrieval-augmented generation systems.
Day-to-day: Chunking strategies, embedding models, vector DB tuning, hybrid retrieval, reranking, query transformation, context window optimization.
Differentiator: Vs. AI Engineer, explicitly narrower. Vs. Context Engineer, more infrastructure-heavy on the retrieval side.
US comp: $150K to $230K base.
Verdict: Niche and fading. Absorbed into AI, LLM, and Context Engineer titles through 2026.

10. AI Systems Engineer #

Stable

Definition: Designs the full end-to-end architecture of AI systems across models, data, orchestration, and UI.
Comp: $180K to $300K base.
Companies: Palantir, NVIDIA, Scale AI, Anthropic.
Verdict: Closer to senior/staff architect rebadge.

B. Developer-Adjacent AI Titles

Titles for developers whose workflow is shaped by AI, plus memes and predictions clustered around them.

11. Vibe Coder #

Meme

Definition: Originally (Karpathy, Feb 2, 2025) someone who “fully gives in to the vibes,” prompts an AI agent, accepts all diffs, and doesn’t read the code. Now used loosely for any AI-assisted coder.
Verdict: Meme, not a job title. Collins Dictionary Word of the Year 2025. No employer hires for it. Many quietly screen against it.
Origin: Andrej Karpathy, Feb 2, 2025.

12. Vibe Engineer #

Workflow Label

Definition: Simon Willison’s Oct 7, 2025 counter-term for seasoned engineers who use AI agents responsibly, with tests, reviews, and accountability.
Verdict: A workflow, not a title. Willison himself called the name “stupid.” Useful as internal shorthand for “AI-augmented senior engineer.”

13. AI-Assisted / AI-Augmented / AI-Native Developer #

Skill, Not Title

Definition: A developer whose workflow is built around AI coding agents (Claude Code, Cursor, Copilot, Codex CLI, Gemini CLI, Windsurf).
Comp: Standard SWE bands.
Verdict: Not a title, a skill. Within 18 months this will simply be “Software Engineer.”

14. Forward-Deployed Engineer (FDE / FDSE) #

Hot but Oversold

Definition: Customer-embedded engineer who deploys complex AI and data platforms into enterprise environments end-to-end.
Day-to-day: Discovery calls with customer executives, requirements translation, building custom integrations on top of the product, training customer teams, iterating on prototypes.
Stack: Python, TypeScript, SQL, Spark, AWS/GCP, Docker/Kubernetes plus the company’s own platform.
Differentiator: Vs. Solutions Engineer, writes real production code. Vs. Applied AI Engineer, much more customer-facing and less internal platform work.
US comp: Palantir average TC $238K, range $205K to $486K, Staff $630K+. OpenAI/Anthropic mid-to-senior $350K to $550K TC. Range overall $180K to $700K+.
Companies: Palantir (origin), OpenAI, Anthropic, Cohere, Databricks, Scale AI, Ramp, Adobe.
Growth: Job postings up over 800% Jan to Sept 2025 per FT. Over 10x YoY on Indeed per WSJ data cited by Gergely Orosz. Public-company earnings-call mentions went from 8 to 50 in a year.
Verdict: Real, hot, and potentially oversold. Orosz’s reporting shows developers quitting after weeks when they realize the role is consulting. Hire carefully. Set expectations brutally.
Origin: Palantir, ~2010s. Went AI-industry mainstream in 2025.

15. “Builder” #

Prediction

Definition: Boris Cherny (Head of Claude Code, Anthropic) predicted on Lenny’s Podcast that “software engineer” as a title will disappear and be replaced by “builder” by end of 2026. Someone who uses AI to ship end-to-end, regardless of specialty.
Verdict: Prediction, not yet a hiring title. A handful of startups have listed “Founding Builder” roles. Zero Fortune 500 orgs have adopted it.

C. Traditional ML Roles

The older discipline. Included for differentiation from AI Engineer titles.

16. Machine Learning Engineer #

Anchor Role

Definition: Builds, trains, and deploys ML models from data to production.
Stack: Python, PyTorch/TensorFlow, XGBoost, Spark, MLflow, Kubeflow, feature stores.
Differentiator: Vs. AI Engineer, actually trains models. Vs. Data Scientist, owns production.
US comp: Levels.fyi median $265,000 TC.
EU: UK median £105K base. Zurich and Dublin €140K to €220K.
Verdict: The stable anchor role. Not going anywhere.

17. ML Research Engineer #

Elite

Definition: Engineer embedded in research teams. Implements papers, runs experiments, builds training infrastructure.
US comp: $300K to $700K TC at frontier labs (OpenAI, Anthropic, DeepMind, Meta FAIR). Can hit $1.4M at the top.
Verdict: Stable, rare, elite.

18. Data Scientist (AI-focused) #

Declining

US comp: $140K to $240K TC.
Verdict: Declining in pure form. Many rebranding to AI Engineer or Applied Scientist.

19. MLOps Engineer #

Stable

Definition: Productionizes and maintains ML pipelines, monitoring, retraining, infrastructure.
US comp: Glassdoor median $161,317 (range $132K to $199K). Senior/Lead $160K to $220K+ base.
Verdict: Stable. LinkedIn Emerging Jobs noted 9.8x 5-year growth.

20. AI Research Scientist / Applied AI Scientist #

Elite, PhD

Definition: PhD-track role. Publishes papers and/or drives novel model work.
US comp: $400K to $1.5M+ TC at frontier labs.
Verdict: Stable, elite, PhD gatekept.

21. Research Engineer (at frontier labs) #

Stable

Definition: Effectively the unified title at Anthropic, OpenAI, and DeepMind covering what others call ML Research Engineer, Applied Scientist, or Alignment Engineer. Anthropic explicitly notes its “engineers do lots of research, researchers do lots of engineering” and both author papers.
US comp: $350K to $1.4M+ TC.
Verdict: The dominant title inside AI labs.

D. Infrastructure & Ops AI Titles

The plumbing. Compute, platforms, monitoring, reliability. Where the money actually goes.

22. AI Infrastructure Engineer #

Hot

Definition: Builds and maintains GPU clusters, model-serving infra, distributed training stacks.
Stack: CUDA, Triton, Ray, vLLM, SLURM, Kubernetes, Terraform, NVIDIA GB200/H200.
US comp: $220K to $600K TC. Over $1M at NVIDIA, Anthropic, OpenAI at staff+.
Verdict: Hot and stable. Compute is the bottleneck.

23. AI Platform Engineer #

Growing

Definition: Owns the internal AI platform (eval harnesses, model registry, orchestration, observability) that AI Engineers build on top of.
US comp: $180K to $400K TC.
Verdict: Stable, growing. Most large orgs now have a dedicated AI Platform team.

24. AI Ops / AIOps Engineer #

Naming Collision

Definition: Two meanings. Legacy: IT operations augmented by AI for incident detection (Moogsoft, Dynatrace). Emerging: Ops for autonomous agent systems (AgentOps).
US comp: $140K to $210K base.
Verdict: Naming collision. Disambiguate in every JD.

25. LLMOps Engineer #

Stable

Definition: MLOps for LLMs. Prompt versioning, eval pipelines, RAG monitoring, model routing, token-cost observability, hallucination metrics.
Differentiator: Vs. MLOps, qualitative evals, hallucinations, prompt and RAG versioning vs. traditional drift metrics.
US comp: $160K to $240K base. Often 10 to 15% premium over MLOps.
Verdict: Stable subspecialty.

26. AI Reliability Engineer (AIRE) #

Emerging

Definition: SRE-style role for AI agent uptime, guardrails, and failure recovery.
US comp: $180K to $280K base.
Verdict: Emerging. Stable trajectory.

27. Model Deployment Engineer #

Stable

Definition: Serves and optimizes models in production (quantization, batching, autoscaling).
US comp: $160K to $260K base.
Verdict: Often subsumed under AI Infrastructure.

E. Product & Business AI Titles

Roles outside engineering that now carry the “AI” prefix. Includes the C-suite.

28. AI Product Manager #

Stable & Growing

Definition: PM for AI-powered products. Responsible for eval design, model and feature trade-offs, cost/quality/latency triangle, safety posture.
US comp: ZipRecruiter average $159,405. Base $130K to $200K, total comp $180K to $260K+. Top packages over $300K at leading tech cos.
Verdict: Every product org needs one.

29. AI Solutions Architect #

Stable

Definition: Pre-sales and post-sales architect designing customer AI solutions.
US comp: $170K to $320K base plus variable comp.
Verdict: Mirrors the cloud solutions architect pattern.

30. AI Strategist #

Squishy

Definition: Advisory role bridging C-suite and engineering on AI adoption roadmaps.
US comp: $140K to $280K base.
Verdict: Squishy title. Often a consulting rebrand.

31. Chief AI Officer (CAIO) #

Hot

Definition: C-suite executive owning AI strategy, governance, portfolio, and risk across the organization.
US comp: Glassdoor average $352,438. Top earners over $500K. Enterprise ($1B+ rev) base $350K to $450K. Big tech $400K to $500K+ base with RSUs taking TC to $1M to $2M. Comparably average $259K.
EU: UK £180K to £320K base. Germany €170K to €300K. Nordics €150K to €270K base.
Growth: CAIO postings up ~400% since 2023. About 60% of global orgs now have a dedicated AI executive.
Driver: EU AI Act, White House EO requiring federal agencies to appoint CAIOs (Oct 2023), regulatory gravity.
Companies: Dell, NASA, ODNI, Mastercard, JPMorgan, Accenture, UPS, Visa, US DoD.
Verdict: New but real C-suite role. Likely consolidates with CTO long-term (5 to 7 years), but hot through 2027.

32. Head of AI #

Stable

Definition: Senior leader of AI function. Often one level below CAIO, or equivalent at smaller orgs.
US comp: $250K to $500K base plus equity.
Growth: “Head of AI” jobs up 3x over five years per LinkedIn.
Verdict: Stable.

F. Specialized & Emerging Titles

Niche but real roles: trainers, red teamers, evals, safety, ethics, and model behavior.

33. AI Trainer / RLHF Annotator #

High Volume

Definition: Produces high-quality human feedback data to train and align models. Often domain experts (lawyers, doctors, coders).
US comp: $30 to $200/hour contractor. FTE $80K to $200K. Domain experts at Scale, Surge, Mercor can bill $100 to $300/hr.
Verdict: Huge volume, low prestige, contractor-heavy. Real work.

34. AI Red Teamer #

Growing

Definition: Adversarial tester for AI models and systems. Jailbreaks, prompt injection, safety/bias failures, cyber capabilities.
US comp: Entry $60K to $90K (10a Labs). Mid $120K to $160K base. Senior $200K to $400K at frontier labs.
Companies: Anthropic’s Frontier Red Team, OpenAI Preparedness, Microsoft AI Red Team, Google, NVIDIA, HiddenLayer, 10a Labs, Mindgard.
Origin: OpenAI GPT-4 system card, 2023.
Verdict: Stable, growing under EU AI Act and NIST AI RMF.

35. Evals Engineer #

Undervalued

Definition: Designs and runs evaluation harnesses for LLM quality, safety, and behavior.
Stack: LangSmith, Braintrust, Inspect, Weights & Biases, custom harnesses.
US comp: $160K to $280K base. $400K+ TC at frontier labs.
Verdict: The most under-valued title relative to its importance. Growing fast.

36. AI Alignment / AI Safety Engineer #

Prestigious

Definition: Works on RLHF, Constitutional AI, interpretability, model steering, policy enforcement.
US comp: $300K to $900K+ TC at Anthropic, OpenAI, DeepMind.
Companies: Anthropic (Alignment team), OpenAI Safety Systems and Preparedness, Google DeepMind Safety, Apollo Research, METR, Redwood Research.
Verdict: Real, prestigious, narrow.

37. AI Ethics Officer #

Often Absorbed

Definition: Governance and compliance role for fair, auditable, regulation-compliant AI.
US comp: $140K to $280K base.
Verdict: Stable but often rolled into CAIO office or Legal.

38. Model Behavior Engineer #

Niche Elite

Definition: Shapes how models respond. Persona, tone, refusal policies, house style. OpenAI’s Model Behavior team is the canonical example.
US comp: $250K to $700K TC.
Companies: OpenAI (explicit team), Anthropic (Claude character work), Google.
Verdict: Niche, elite, stable at frontier labs only.

G. Turkovic Predictions (2026 to 2028)

Two roles that don’t exist at scale yet. I expect them to be the most important new engineering titles by 2028. Named here so companies adopting them have shared vocabulary.

39. AI Delivery Engineer Turkovic Prediction #

Predicted

Definition

The human owner of the arc from vague business intent to shipped outcome, using AI as the primary production tool. Not a project manager. Not a prompt engineer. The person who turns “we need X that handles these edge cases” into a working thing, with the AI doing most of the typing and the human doing all of the thinking.

Day-to-day

Refines story descriptions into precise specifications. Decides which AI tools, models, and automation levels are appropriate for each task (high-stakes code gets heavy guardrails; low-stakes code gets auto-accepted). Drives AI agents through generation, review, and iteration. Owns the final shipped increment.

Differentiator

Vs. AI Engineer: AI Engineer builds AI features into a product. AI Delivery Engineer uses AI to build the product itself.
Vs. Tech Lead: Tech Leads coordinate human engineers. AI Delivery Engineers coordinate AI agents and humans.
Vs. Product Manager: PMs define what to build. AI Delivery Engineers ship it.
Vs. “Builder” (Cherny prediction): “Builder” is a vibe label. AI Delivery Engineer is a concrete role with defined outputs.

Stack

Claude Code, Cursor, Codex CLI, Windsurf, Devin-class agents. Spec-driven dev tools (GitHub Spec Kit, Amazon Kiro). Eval harnesses. Versioned prompt libraries. Observability for agent runs.

Seniority

Junior → Mid → Senior → Staff. Expect Staff AI Delivery Engineers to run 5 to 10 agent fleets in parallel.

Comp (projected)

Bands track Senior Software Engineer and Tech Lead bands. $180K to $350K TC for senior in US; €95K to €170K in EU. Premium of 10 to 15% over equivalent non-AI SWE for the same seniority.

Already forming

Shopify (post-Lütke memo), Anthropic internal eng, Cursor, Vercel, most YC W25/W26 companies under 20 heads.

Verdict

The most common engineering title in companies under 200 people by 2027. In larger companies it will exist alongside traditional SWE and AI Engineer titles. In smaller companies it will replace most of them.

40. Verification Engineer Turkovic Prediction #

Predicted

Definition

QA tester fused with domain expert. Owns both end-to-end testing of AI-produced systems and verification that the AI built the right thing for this specific business. Maintains the domain knowledge base that makes correct verification possible.

Day-to-day

Runs the system end-to-end against real and synthetic scenarios. Hunts broken edge cases. Maintains and extends a domain-specific knowledge base (regulatory rules, business-logic exceptions, industry conventions, product-specific invariants). Signs off on AI-produced code before it ships to anything touching money, safety, health, or compliance. Designs and runs eval suites tied to business outcomes, not just code correctness.

Differentiator

Vs. QA Engineer: QA tests that the code works. Verification Engineer tests that the code does the right thing for the domain.
Vs. Evals Engineer: Evals Engineer measures model quality. Verification Engineer measures system correctness for this business.
Vs. AI Safety Engineer: Safety Engineer prevents harmful outputs across all users. Verification Engineer prevents wrong outputs for this specific organization’s business rules.
Vs. AI Red Teamer: Red Teamer adversarially tests for worst-case behavior. Verification Engineer verifies expected behavior for the domain.

Stack

Property-based testing, scenario-based eval frameworks, domain knowledge bases (often graph-based or structured), internal compliance databases, observability tools, human-in-the-loop review systems.

Seniority

Mid → Senior → Staff → Principal. Senior+ will typically carry formal domain credentials (CPA, JD, MD, series licenses, industry certifications).

Comp (projected)

Premium over traditional QA roles. $140K to $260K base in US; €80K to €150K in EU. In regulated industries (fintech, healthcare, pharma, defense) staff+ can reach $300K to $450K TC.

When it fits

Large organizations (200+ engineers) where ownership of AI output is diffuse.
High-stakes domains: fintech, healthcare, legal, industrial control, compliance-heavy regulated spaces.
Any code path touching money, safety, regulatory exposure, or critical business logic.

When it doesn’t

Low-stakes code, internal tooling, experiments. For these, AI self-checks or auto-accept workflows will suffice.

Verdict

The quiet gatekeeper role of 2027 to 2028. Not glamorous. Not LinkedIn-famous. But the role that lets regulated businesses actually deploy AI-generated systems without eating an incident every quarter. Expect domain-specific variants: “Verification Engineer, Payments” or “Verification Engineer, Clinical Workflows.”

No roles match your search. Try a broader term or clear the filter.

H. Differentiation Matrix

When two titles look similar, this is the question you’re actually asking.

If you need…	You want	Not
Someone who trains models from scratch	ML Engineer / Research Engineer	AI Engineer
Someone who calls pretrained APIs	AI Engineer / LLM Engineer	ML Engineer
Focus on prompt phrasing only	Prompt Engineer (declining)	Context Engineer
Focus on retrieval, memory, tools	Context Engineer	Prompt Engineer
Multi-step autonomous agents	Agent / Agentic AI Engineer	RAG Engineer
GPU cluster plus serving stack	AI Infrastructure Engineer	AI Platform Engineer
Internal dev-facing AI platform	AI Platform Engineer	AI Infrastructure Engineer
Monitor LLMs in production	LLMOps Engineer	MLOps Engineer
Embedded with enterprise customer	Forward-Deployed Engineer	Applied AI Engineer
Adversarial testing	AI Red Teamer	AI Safety Engineer
Eval harness design	Evals Engineer	Prompt Engineer
RLHF, Constitutional AI	Alignment Engineer	AI Safety Engineer (overlap)
Shape tone, persona, refusals	Model Behavior Engineer	Alignment Engineer
Own AI strategy for whole company	Chief AI Officer	Head of AI (sub-scope)
Own arc from intent to shipped outcome using AI	AI Delivery Engineer	Tech Lead
Verify AI-built systems match domain business logic	Verification Engineer	QA Engineer

I. Growing vs Declining (2025 to 2026)

Which titles are actually hiring, which are converging, and which exist mostly on blog posts.

Growing sharply

Forward-Deployed Engineer: +800% postings Jan to Sept 2025.
Context Engineer: near-zero to 1,000 at Cognizant alone.
Chief AI Officer: +400% since 2023. Head of AI +3x in 5 years.
Agent / Agentic AI Engineer: +986% YoY.
AI Engineer: “fastest-growing role in the US.”
Red Teamer, Evals, Safety: growing off small base due to EU AI Act.

Plateauing

Prompt Engineer: absorbed into AI Engineer and Context Engineer.
GenAI Engineer: consolidating under AI Engineer.
RAG Engineer: becoming a skill, not a job title.

Declining

Data Scientist (AI-focused): rebranding to AI Engineer or Applied Scientist.
AIOps in its legacy IT-ops sense: getting name-collided by agent ops.

Predicted emerging

AI Delivery Engineer (Turkovic, 2026): dominant in sub-200 companies by 2027.
Verification Engineer (Turkovic, 2026): standard in regulated enterprises by 2027 to 2028.
“Builder” (Cherny, Anthropic): predicted to replace “software engineer” by end of 2026.
“Specification Engineer” (Grove, OpenAI): predicted from the thesis that specifications supplant code.

J. Recommended Standardization for HR

Stop paying 20 to 40% premiums for title arbitrage. Collapse the taxonomy.

Five engineering buckets

AI Engineer (includes Applied AI, GenAI, LLM, RAG, Agent, Prompt, Context variants)
Machine Learning Engineer (includes ML Research Engineer)
AI Platform / Infrastructure Engineer (includes MLOps, LLMOps, Model Deployment, AI Reliability)
Research Engineer / Research Scientist (frontier and applied)
Forward-Deployed / AI Solutions Engineer (customer-embedded)

Three specialist tracks

AI Safety / Red Team / Evals
AI Product Manager
Chief AI Officer / Head of AI

Two emerging tracks to prepare for 2027

AI Delivery Engineer (end-to-end AI-assisted delivery ownership)
Verification Engineer (domain-specific verification of AI-produced systems)

Everything else is seasoning. Put specialization in the JD, not the title.

Final Words

This reference is a living document. I’ll revise it as the market moves. If your company is using a title I didn’t cover, or if you think I got the comp range wrong, let me know.

Find me on LinkedIn (linkedin.com/in/ivanturkovic), X (x.com/ithora), and Threads (threads.com/@ithora). For consulting engagements or fractional CTO work, reach out at ivanturkovic.com/contact or book directly at cal.eu/ivan-turkovic/30min.

If you’re an HR leader using this guide to standardize roles at your company, what title did you collapse and what did you keep? I want to hear how this plays out in the wild.