New opportunity

AI and Machine Learning Engineering Graduate

HPE · Durham, United States of America

About this role

HPE seeks an AI and Machine Learning Engineering Graduate to develop models and software algorithms that use structured and unstructured data in products and systems. This full-time role is based in Durham, North Carolina, and is designated on-site, with work primarily from an HPE office. The schedule is 40 hours per week.

Responsibilities

Design, develop and implement AI and machine learning models, including data preprocessing, feature engineering, algorithm selection, training and evaluation. Gather and analyze datasets, apply statistical methods and visualize results to inform decisions.
Improve model performance through experiments, hyperparameter tuning, optimization, feature selection and dimensionality reduction. Translate technical specifications, customer requirements and industry trends into AI/ML solutions and systems improvements; test and debug code, document procedures and monitor performance metrics.
Collaborate with data scientists, software engineers, product managers and domain experts to align requirements and integrate solutions into products or existing systems. Participate in design reviews and team stand-ups, incorporate feedback, maintain project documentation and present findings and recommendations to stakeholders. The role may also involve reporting data findings and maintaining business intelligence tools, databases or dashboards.

Qualifications

A bachelor’s degree in computer science, engineering, data science, machine learning, artificial intelligence or a closely related quantitative discipline is required. The posting specifies typically 0–2 years of experience; a master’s degree is desirable.
Proficiency in a programming language such as Python, R or Java; understanding of statistics, probability, linear algebra, calculus and optimization; and knowledge of supervised and unsupervised learning, deep learning, neural networks, reinforcement learning and natural language processing.
Ability to work with large datasets, clean and preprocess data, and perform exploratory analysis. Experience with SQL, databases and visualization tools such as Matplotlib or Tableau is required. Knowledge of TensorFlow, PyTorch or scikit-learn is highly beneficial.

Compensation and benefits: The North Carolina annual base salary range is USD 81,700–187,800; actual offers may vary, and variable incentives may also be offered. HPE describes health and wellbeing benefits and professional development programs.

For full details about this role, click “Apply on company site” to view the employer’s original posting.

Skills for this role

PythonRJavaTensorFlowPyTorchscikit-learnSQLMatplotlibTableauMachine LearningDeep LearningStatistical ModelingData PreprocessingData CleaningFeature EngineeringFeature SelectionModel TrainingModel EvaluationHyperparameter TuningDimensionality ReductionData VisualizationExploratory Data AnalysisSupervised LearningUnsupervised LearningNeural NetworksReinforcement LearningNatural Language ProcessingStatisticsProbabilityLinear Algebra

Checking your profile…

Preparation suggestions, not the employer’s actual interview questions. Always check the original posting for current requirements.

Understand the role

What matters most

Build and evaluate models end to end
The role covers preprocessing, feature engineering, algorithm selection, training and evaluation. Prepare to explain the reasoning behind each step in a model-development workflow.
Turn data into reliable evidence
The posting calls for gathering and analyzing datasets, statistical methods and visualization, with SQL and database experience required. Prepare to show how data cleaning and exploratory analysis inform modeling decisions.
Improve performance through experiments
Responsibilities include hyperparameter tuning, optimization, feature selection and dimensionality reduction. Prepare to explain how you would compare experiments and decide whether a change improves the chosen metric.
Integrate and communicate work within a team
The role involves translating requirements into AI/ML solutions, integrating them into products or systems, and documenting and presenting results. It also calls for design reviews and incorporating feedback; prepare to explain technical choices clearly and revise them against project goals.

Likely challenges to think through

Working across varied data
A likely difficulty is choosing useful preprocessing for large structured or unstructured datasets without losing important information. Prepare to discuss how you would inspect data quality, establish a baseline and check the effect of cleaning decisions; the posting does not identify a specific dataset.
Balancing model gains with system needs
A more complex model or extensive tuning may improve an evaluation score without being the best fit for an existing product or system. Prepare to explain how you would weigh measured performance against implementation and maintenance considerations; no specific product constraints are stated.
Making sound decisions within defined scope
This graduate role describes independent judgment within defined parameters, alongside manager or team-lead feedback. A likely challenge is knowing when to proceed with an experiment and when to seek alignment. Prepare an example of presenting evidence, uncertainties and a recommendation for review.

Practice your answers

Technical practice

How would you use SQL and exploratory analysis to assess a dataset before modeling?

What to coverThe posting calls for SQL, database experience, exploratory analysis, and visualization. Suggested answer focus: explain how you would inspect distributions and missing values, check whether the data suits the objective, and communicate findings with a clear chart.

How would you prepare structured and unstructured data for a model?

What to coverThe role involves both data types and explicitly includes cleaning and preprocessing. Suggested answer focus: discuss data-quality checks, appropriate transformations for each type, and how to keep training and evaluation data separate.

How would you choose between supervised, unsupervised, and deep-learning approaches?

What to coverThe posting names all three areas and assigns algorithm selection to the role. Suggested answer focus: relate the choice to available labels, the intended outcome, data volume, complexity, and the value of a simpler baseline.

How do probability, linear algebra, calculus, and optimization inform model training?

What to coverThe posting identifies these foundations as important for building and evaluating models. Suggested answer focus: connect uncertainty, vector operations, gradients, and objective functions to a concrete training example without assuming a particular framework.

How would you evaluate whether a model improves a product or system?

What to coverThe role includes model evaluation and monitoring performance metrics to improve quality and accuracy. Suggested answer focus: define a baseline and metrics tied to the objective, use an appropriate test set, and distinguish a meaningful improvement from an apparent one.

How would you design a hyperparameter-tuning experiment?

What to coverThe posting explicitly mentions experiments, hyperparameter tuning, and performance optimization. Suggested answer focus: describe a controlled comparison, consistent evaluation criteria, a held-out test set, and the trade-off between improvement and computation.

When would you use feature engineering, feature selection, or dimensionality reduction?

What to coverAll three are named in the responsibilities. Suggested answer focus: distinguish creating useful inputs from removing or compressing them, and explain how you would check whether a change helps rather than discarding important information.

What would you consider when building a natural-language processing model?

What to coverNatural language processing is listed among the expected knowledge areas, and the role handles unstructured data. Suggested answer focus: discuss text preparation, choosing a baseline and evaluation measure for the task, and common sources of misleading results.

How does reinforcement learning differ from supervised learning, and when might it be inappropriate?

What to coverThe posting lists both as knowledge areas but does not specify a reinforcement-learning project. Suggested answer focus: contrast reward-driven interaction with learning from labeled examples, and explain when feedback or experimentation constraints favor a simpler approach.

How would you test, debug, and monitor an ML component integrated into an existing system?

What to coverThe job family includes testing, debugging, documentation, and performance monitoring; the responsibilities include integration with existing products or systems. Suggested answer focus: describe checks for data and code behavior, reproducible documentation, and metrics that could reveal degradation after integration.

Behavioural practice

Tell me about a time you worked with people in different roles to clarify a technical requirement.

What to coverThe posting emphasizes collaboration with data scientists, software engineers, product managers, and domain experts. Suggested answer focus: use a real situation, the questions and actions you took to align objectives, and the resulting decision or outcome.

Describe a time you incorporated critical feedback on a design or implementation.

What to coverThe role includes design reviews and actively incorporating feedback from a manager or team leader. Suggested answer focus: explain the situation, what you changed or respectfully challenged, and the result.

Tell me about a time you raised a blocker and coordinated next steps with teammates.

What to coverThe posting mentions stand-ups for progress updates, challenges, and coordination. Suggested answer focus: give the situation, how you communicated the blocker and acted within your scope, and what happened next.

Describe a time you explained a technical finding to a stakeholder.

What to coverThe responsibilities include reports, presentations, and recommendations to stakeholders. Suggested answer focus: identify the audience and situation, explain how you made the evidence understandable, and state the resulting decision or feedback.

Tell me about a time you made a technical judgment with limited scope or guidance.

What to coverThe graduate-level description calls for independent judgment within defined parameters and recommendations that support team efforts. Suggested answer focus: describe the boundaries, how you evaluated options or sought input, the action you took, and its result.

Sharpen your knowledge

Topics to prepare

Mathematics behind model decisions
The posting calls for statistics, probability, linear algebra, calculus and optimization. Suggested exercise: explain how a loss function, gradient and probability estimate affect training and evaluation of a simple model.
Supervised and unsupervised learning
The posting names both approaches and requires algorithm selection. Suggested exercise: compare a labeled prediction problem with an unlabeled grouping problem, explaining the algorithm choice and how each result could be evaluated.
Deep learning and other named approaches
The posting names neural networks, deep learning, reinforcement learning and natural language processing. Suggested exercise: outline what data and objective each approach needs, then explain which would suit a text-analysis task and why. This is conceptual revision, not a request to build every type of model.
Model evaluation and statistical interpretation
The role involves statistical analysis, training and evaluation, and monitoring performance metrics. Suggested exercise: choose a metric for a hypothetical prediction task, explain what it misses, and describe how you would interpret a change in results.
Performance-improvement methods
The posting explicitly lists hyperparameter tuning, experiments, optimization, feature selection and dimensionality reduction. Suggested exercise: explain when you would try each method and how you would decide whether it improved a model.
Requirements and system integration
The role translates customer requirements and technical specifications into AI/ML solutions that fit products or existing systems. Suggested exercise: turn a hypothetical requirement into a model objective, a measurable success criterion and two questions for a design review.

Skills to revise

Program a small ML workflow
Proficiency in a language such as Python, R or Java is required. Suggested exercise: use one of those languages to write and debug code that reads data, trains a model and reports an evaluation result. TensorFlow, PyTorch and scikit-learn are named as beneficial, not required.
Query data with SQL
Experience with SQL and databases is required. Suggested exercise: write queries that join two tables, filter incomplete records and aggregate a field into a dataset suitable for analysis.
Clean and explore data
The posting requires work with large datasets, preprocessing, cleaning and exploratory analysis. Suggested exercise: inspect a dataset for missing values and inconsistent fields, document cleaning decisions, and create features for a stated prediction objective.
Visualize and explain findings
Experience with visualization tools such as Matplotlib or Tableau is required, and the role reports findings to stakeholders. Suggested exercise: make two plots showing data patterns or model results, then give a short explanation of what decisions they support and what remains uncertain.
Run a controlled model comparison
The role includes model training, evaluation, experimentation and hyperparameter tuning. Suggested exercise: compare a baseline with one tuned version using the same evaluation method; record settings, results and the reason for keeping or rejecting the change.
Test, document and present work
The posting calls for testing and debugging code, documenting methods and results, participating in design reviews, and presenting recommendations. Suggested exercise: test a preprocessing step on an edge case, write a brief experiment record, and present one design choice and its trade-off as if seeking team feedback.

Your preparation checklist

Map evidence to the stated qualifications
Prepare a one-page list of relevant coursework or work, code samples, and examples using Python, R, or Java. The posting requires a related bachelor’s degree and programming proficiency; TensorFlow, PyTorch, or scikit-learn knowledge is beneficial, not required.
Build an end-to-end project walkthrough
Choose a real project and prepare a five-minute explanation of its data, preprocessing, feature engineering, model choice, training, evaluation, and limitations. These stages are stated responsibilities; use results you can substantiate.
Show a data analysis artifact
Bring a small SQL example and a chart or dashboard you can explain, including how you cleaned data and what the analysis revealed. The posting calls for SQL, databases, exploratory analysis, and visualization.
Explain an improvement experiment
Prepare a before-and-after results table for an experiment you actually ran, such as hyperparameter tuning or feature selection. State the metric, comparison method, and trade-off; model optimization is part of the role.
Practice a product-integration design exercise
As interview preparation, sketch how an ML model could fit into an existing product: clarify requirements, outline inputs and outputs, and propose tests, debugging steps, documentation, and performance metrics. The posting describes integration and monitoring, but does not specify an interview exercise.
Rehearse a cross-functional update
Prepare a brief example of communicating findings, seeking feedback, or revising a design with teammates. End with a clear recommendation and a concise progress update, reflecting the posting’s collaboration and presentation duties.
Draft questions for the HPE team
Prepare questions about the first assignments, how model performance is evaluated, who reviews designs, and how the Durham on-site team works together. These are questions to ask, not details established by the posting.

Availability last checked: Nov 4, 2026. Status can change after a check; confirm on the employer’s website. Report an issue · Our checking policy

KEEP EXPLORING

Similar AI jobs

Related skills and specializations in United States of America. Matched to this role, not your profile.

Explore more jobs

Finding similar opportunities…