The titles overlap, every company defines them differently, and job descriptions mix them freely. Here is a practical way to tell these roles apart — and to decide which one fits you as a software developer.
The quick version
- ML Engineer — trains, tunes, and serves machine learning models. Heavier on data, math, and model lifecycle.
- AI Engineer — builds applications on top of existing models (mostly LLMs). Heavier on software engineering, RAG, agents, and production.
- GenAI Developer — an AI engineer focused specifically on generative models and LLM-powered features.
- Agentic AI Engineer — an AI engineer specializing in agents: tool use, planning, memory, and multi-step workflows.
- Forward Deployed Engineer (FDE) — an engineer who sits with customers, scopes real problems, and ships tailored AI solutions fast.
Where the lines actually fall
The clearest split is who builds the model vs who builds with the model.
ML engineers live closer to training data, feature pipelines, and model serving. AI engineers and GenAI developers live closer to product code: prompts, retrieval, tools, evaluation, APIs, and deployment. If you are a software developer, the second group is usually the faster, more natural transition.
What each role optimizes for
ML Engineer
Data pipelines, model training and evaluation, MLOps, and serving. Strong Python, statistics, and infrastructure. Great if you enjoy the modeling layer.
AI Engineer / GenAI Developer
Application architecture around LLMs: RAG, function calling, structured outputs, evals, observability, cost and latency, and deployment. This is software engineering with models as components.
Agentic AI Engineer
Everything an AI engineer does, plus the hard parts of autonomy: planning loops, tool orchestration, memory, guardrails, and knowing when a workflow beats a fully autonomous agent.
Forward Deployed Engineer
Customer-facing delivery. You combine AI engineering with discovery, scoping, demos, and communication. High trust, high impact, and increasingly in demand.
Which one should you target?
If you already ship software, aim at AI Engineer / GenAI Developer first, then specialize into Agentic AI or FDE based on what you enjoy:
- Love building product features? AI Engineer / GenAI Developer.
- Love autonomy and orchestration? Agentic AI Engineer.
- Love customers and delivery? Forward Deployed Engineer.
All four share the same core: build real systems, evaluate them, and explain your trade-offs. Start with the AI Engineer Roadmap, then build the projects that match the role you want.