Applied AI / Forward Deployed Engineer · The Role: Applied AI & Forward Deployed Engineer · Lesson 2 of 5
How it differs from ML Engineer, Research & Solutions roles
Interviewers often probe whether you understand the boundaries of the role. Know the neighbors:
- Research Engineer / Research Scientist — improves the model itself (architecture, training, RLHF). Deep ML theory, PyTorch at scale. You are not this.
- ML Engineer (traditional) — trains and serves bespoke models (ranking, fraud, recsys). Feature pipelines, model training, MLOps. Overlaps on eval/serving, but you mostly call a pretrained LLM rather than train one.
- Solutions Architect / Sales Engineer — pre-sales, reference architectures, does less hands-on shipping. FDE is more hands-on and owns delivery of working software.
- Software Engineer (product) — closest cousin. The AAE/FDE is a product SWE who is fluent in LLM behavior, evals, and non-determinism.
Say this in the interview"I sit between the model and the customer. I don't train frontier models — I take one, wrap it in retrieval, tools, evals and guardrails, and ship something reliable against the customer's actual data." Clean role clarity signals seniority.
◆ Lock it in
- Research = improve the model. ML Eng = train bespoke models. You = apply pretrained models to ship product.
- FDE owns delivery — more hands-on than a solutions architect or sales engineer.
- Your differentiator vs a normal SWE: fluency in non-determinism, evals, and LLM failure modes.
Feynman drill — say it out loudIn four sentences, place the AAE/FDE role against its neighbors: research engineer, ML engineer, solutions architect, product SWE.
Step 1 rate your confidence · Step 2 pick your answer
A hiring manager asks how your role differs from a Research Engineer. Best answer?
Step 1 — how sure are you?