Machine Learning Engineer, Specialist

Join the visionary team of Vanguard’s Advice Wealth Management Technology’s Data & Analytics Engineering! We are on the lookout for a passionate Machine Learning Engineer to pioneer innovative models that tackle the most intricate financial planning and portfolio construction challenges. Our AIML group stands at the forefront of technological advancement, boasting the most sophisticated machine learning models driven by cutting-edge methodologies and strategies. Continuously investigating the latest in machine learning technologies, including GenAI, we strive to enhance our solutions and stay ahead of the curve. Embark on this exciting journey with Data & Analytics Engineering and shape the future of Advice Wealth Management Technology! 

This person will participate in end-to-end machine learning projects from conception through deployment and ongoing support. Collaborating with data scientists to dive into complex machine learning challenges that push the boundaries of current technological capabilities, including supervised learning, reinforcement learning, deep learning, and GenAI. In addition, working closely with methodology researchers to integrate valuable insights and strategies to improve investor outcomes. Solve complex problems with multilayered data sets, enhance existing libraries, frameworks and models, and work with data analysts, data engineers, and architecture to identify data distribution differences affecting model performance. 

The successful candidate will: 

• Bring proficiency and excellent understanding of the AWS cloud platform and services, including but not limited to AWS Sage Maker, AWS Lambda, S3 buckets, Step Functions, EMR, Glue, and other services that support building machine learning platforms. 

• Have an excellent understanding of the machine learning development cycle, including data engineering, exploratory data analysis, modeling, and machine learning implementation and operations. 

• Design and implement scalable machine learning solutions, develop predictive models using advanced deep learning and statistical techniques, collaborate with data science and engineering teams to integrate ML solutions, and perform rigorous model evaluation and optimization. 

• Be proficient in software development and well-versed with developer tools such as, but not limited to, Python, VS Code, and Jupiter Notebooks. 

• Be passionate about new advances in machine learning, knowledgeable about supervised learning, reinforcement learning, deep learning, and GenAI. 

• Demonstrate knowledge of AWS security practices, including IAM, S3 bucket policies, security groups, and VPCs. 

• Understand best practices for model training, deployment, and operations, including hyperparameter optimization, model evaluation, and operationalizing ML solutions. 

• Utilize popular Python frameworks such as TensorFlow, PySpark, PyTorch, and Pandas. 

• Leverage software design patterns to develop modular, maintainable, and scalable code. 

Responsibilities:

• Leverages data pipeline designs and supports the development of data pipelines to support model development. Proficient with software tools that develop data pipelines in a distributed computing environment (PySprak, GlueETL).

• Supports integration of model pipelines in a production environment. Develops understanding of SDLC for model production.

• Reviews pipeline designs, makes data model design changes as needed. Documents and reviews design changes with data science teams.

• Supports data discovery & automated ingestion for model development. Performs detailed analysis of raw data sources for data quality, applies business context, and model development needs.

• Engages with internal stakeholders to understand and probe business processes in order to develop hypotheses. Brings structure to requests and translates requirements into an analytic approach. Participates in and influences ongoing business planning and departmental prioritization activities.

• Runs model monitoring scripts, follows process for alerts to management as needed. Addresses issues found in data pipelines from model monitoring