Machine Learning Engineer III

As a Machine Learning Engineer III at Wizard, key responsibilities include building and maintaining training pipelines for fine-tuning open-source LLMs, deploying and serving models on GPU configurations, addressing customer issues related to model shortcomings, and staying updated with the latest in the generative AI field to enhance our systems.

To qualify for the Machine Learning Engineer III position at Wizard, candidates should hold an MS/Ph.D. in computer science, mathematics, or a related quantitative field, possess 1-3 years of ML experience, and demonstrate proficiency in developing ML-driven products at scale. Familiarity with PyTorch, data systems, and frameworks such as Flask or FastAPI is also essential.

A Machine Learning Engineer III at Wizard will work on projects related to dialog systems, which include fine-tuning LLMs, implementing data pipelines, and creating solutions for novel product issues arising from ML challenges, all aimed at enhancing the user shopping experience.

Wizard promotes continuous learning and growth for its Machine Learning Engineer III team members through exposure to the latest developments in generative AI, collaborative problem-solving, and a fast-paced environment where innovation is encouraged.

The expected salary for a Machine Learning Engineer III position at Wizard is around $185,000, which can vary based on individual skills, experience, and qualifications, reflecting our commitment to competitive compensation.

When answering this question, provide specific examples from your past work, discuss the models you’ve worked with, the datasets used, and the results achieved, focusing on your understanding of the iterative process involved in fine-tuning.

Describe your step-by-step approach to creating a training pipeline, including data collection, preprocessing, model training, validation, and monitoring, emphasizing your understanding of best practices and tools like Airflow or Kubernetes.

Share your methodology for diagnosing problems, such as analyzing model output, seeking customer feedback, and working with cross-functional teams to identify and implement solutions, highlighting examples of past successful resolutions.

Discuss any experience you have with deploying models using frameworks like Flask or FastAPI. Be sure to mention challenges you've overcome in production, such as optimizing for runtime performance and ensuring reliability.

Mention resources such as key journals, conferences, or online courses you follow to keep your knowledge current. You can also highlight participation in relevant communities, discussions, or projects that aid your learning.

Share a specific example highlighting the problem, how you approached the iteration process, what changes you implemented, and the impact of your adjustments on the final outcome.

Explain your method for assessing performance bottlenecks using quantitative metrics, and detail how you evaluate trade-offs between model complexity and computational resource costs, supported by previous experiences.

Identify the programming languages and tools you're proficient in, such as Python and PyTorch, and discuss how you have leveraged them in past projects to deliver effective machine learning solutions.

Provide insights into how you've structured and managed data pipelines in past roles, emphasizing challenges you've faced, tools employed (like Pandas), and how these systems support machine learning workflows.

Discuss approaches like cross-validation, regularization techniques, and continuous monitoring of model performance post-deployment to ensure the longevity and reliability of your ML solutions.