Machine Learning Engineer Job Description
A machine learning engineer job description template: the ATS keywords recruiters screen for, a realistic salary range, and how the title differs from AI Engineer.
"Machine Learning Engineer" is usually the more conservative of the two AI-adjacent titles that boomed since 2023. Where "AI Engineer" often names work built on top of someone else's model, this one still means building, training, evaluating and serving your own. If a posting titled Machine Learning Engineer spends most of its requirements on calling a vendor's API rather than on a training pipeline, either the title is being used loosely or the role has quietly become something else.
The other confusion worth clearing up before you apply: Machine Learning Engineer sits between Data Scientist and Data Engineer on most teams, and the boundary moves with company size. At a smaller company, one person may do all three — explore the data, build the model, and deploy it. At a larger one, a Data Scientist prototypes in a notebook, a Data Engineer owns the pipelines feeding it, and the Machine Learning Engineer is the one who has to make the model survive contact with production traffic, changing data, and a pager.
The seniority bar here is less about publications than about scars. A senior candidate is expected to have watched a model degrade in production and know why — data drift, a silent upstream schema change, a feature that was available at training time but not at inference time — not just to have shipped a high-accuracy model in a notebook. That difference rarely shows up on a resume's skills list; it shows up in how someone talks about a project that went wrong.
Bellamyre · Chicago, IL
Full-time · On-site
$150,000 – $195,000
About the role
Bellamyre builds forecasting and inventory-planning software that mid-market retailers and distributors use to plan demand months out. We're hiring a senior Machine Learning Engineer to take the forecasting and anomaly-detection models our data science team builds and turn them into services the product can depend on, day after day.
This is a build-and-operate role, not a research role. The models already exist; the job is training pipelines that don't silently break, serving infrastructure that holds up under real customer traffic, and enough monitoring that we know a model has degraded before a customer's forecast does. You'll work on-site with the Applied ML team in our Chicago office, alongside data engineering and data science.
What you'll do
- Take models from prototype notebooks through production-grade training pipelines and deployment.
- Build and maintain model-serving infrastructure — batch and real-time inference paths, latency budgets, autoscaling.
- Own the feature pipelines and training data that feed production models, in close coordination with data engineering.
- Monitor deployed models for data drift and prediction-quality regressions, and drive retraining when the metrics call for it.
- Run structured batch evaluation and staged rollouts — shadow traffic, canaries — before a model reaches production traffic.
- Partner with data scientists on model design and with backend engineers on the APIs that expose predictions to the product.
- Manage compute cost and GPU utilization for training and serving, since inference cost scales with usage in a way a notebook never has to.
- Document model behavior, known limitations and failure modes for other engineers and for compliance review.
What we're looking for
- Five or more years building and shipping machine learning systems into production, not only research or prototyping.
- Strong Python and hands-on experience with a training framework — PyTorch or TensorFlow.
- Practical experience across the full ML lifecycle: data preparation, training, evaluation, deployment and monitoring.
- Comfort with SQL and large-scale data processing, such as Spark, for building training datasets.
- Working knowledge of containerization and cloud infrastructure — Docker and Kubernetes — for deploying models as services.
- Ability to explain a model's tradeoffs — accuracy, latency, cost — to stakeholders who are not ML specialists.
Nice to have
- Experience with feature stores or real-time feature pipelines.
- Familiarity with classical ML — gradient boosting, time-series forecasting — alongside deep learning; not every production model is a neural network.
- Exposure to model monitoring and observability tooling, whether a commercial product or something built in-house.
- Experience running A/B tests or another method of measuring a model's real impact on a business metric, not just its validation accuracy.
- Prior on-call ownership of a production ML system.
- An advanced degree in a quantitative field — common on this team, though not required.
Benefits
- Medical, dental and vision coverage, with employee premiums covered in full.
- 401(k) with a company match, vested immediately.
- On-site role in Bellamyre's Chicago office, with relocation assistance for candidates moving from out of state.
- $2,000 annual learning budget, usable on conferences, courses or books.
- Twenty days of paid time off plus company holidays.
Salary range
As posted for this sample role. Real pay varies by employer, location and experience.
$150,000–$195,000/ yr
ATS keywords for this role
The applicant tracking system (ATS) — the recruiting software a hiring team searches and filters applicants with — will screen for these. Weight shows how central each one is to this specific posting.
Required and central (4)
Important (8)
Mentioned in passing (7)
Frequently asked questions
What's the difference between a Machine Learning Engineer and a Data Scientist?
In practice, a lot of overlap — but where a distinction holds, the Data Scientist explores the data and builds the first working model, and the Machine Learning Engineer makes it survive production: real traffic, changing data, and a pager. At smaller companies, one person often does both.
Is Machine Learning Engineer the same job as AI Engineer?
Not usually, though the titles get used loosely. Machine Learning Engineer still tends to mean training and serving your own models, including classical ML rather than only deep learning. AI Engineer more often means building features on top of someone else's foundation model. Check the requirements section rather than the title.
Do I need a PhD to work as a Machine Learning Engineer?
No — a PhD matters much more for research-scientist tracks, where the job is publishing new methods. Production machine learning engineering rewards solid software engineering plus applied ML fundamentals more than a doctorate, though an advanced degree is common on teams doing custom modeling.
This was a sample. Your resume should be tailored to the real thing.
Rezi Ninja reads an actual job posting and rewrites your resume to match it, with a Ninja Score so you know it lands before you hit send. Free to start, with the AI usage included.