Software Engineer, Machine Learning Platform

Machine Learning Platform Team

The Machine Learning Platform (MLP) team builds the foundation that every ML and GenAI workload at Wealthsimple runs on. We partner with data science and engineering teams across the company to take models from experimentation to production reliably and at scale. Our responsibilities are to:

• Build standardized, framework-agnostic services for model training, efficient inference, evaluation and monitoring — across both traditional ML and GenAI
• Provide centralized, governed access to internal and external LLMs
• Set the technical direction for how ML and LLM models are trained, shipped and operated across the company

The Role

We are building a multi-product company with a growing offering including Invest, Cash, Crypto, Trade and Tax products.

As a Software Engineer on the Machine Learning Platform team, you will play a pivotal role in our mission to build a self-serve ML Platform to enable data scientists to do their best work. As such, you will work closely with the area owners, data scientists and other software engineers to improve our ML-ops tooling and support the launch of new data products.

In this role, you will have the opportunity to:

• Partner with stakeholders across different product domains, infrastructure to tackle technical problems at scale, which includes building framework agnostic services to support all of Wealthsimple's machine learning needs.
• Establish roadmap and architecture for Wealthsimple's Machine Learning platform
• Build highly performant scalable systems, contributing to our ML platform on Kubernetes and Ray
• Work with a large amount of autonomy to set the technical direction for scalable projects.

We are looking for people who have:

• 5+ years of experience implementing production-ready services
• 3+ years of experience with python
• 5+ years of software development experience in ML infra, ML tooling or ML engineering
• Proven experience designing and implementing highly performant scalable systems with good observability
• Proven experience owning a project end-to-end and owning and tending complex systems for the long run
• Preferred experience with ML ecosystem (e.g. pytorch, transformers, Ray, feature stores, elastic search and similar vector databases, triton, cuda, DVC, vLLM)

Our stack includes:

• Kubernetes to orchestrate our services
• FastAPI for RESTful APIs
• MLflow for experiment tracking
• Kubernetes, Ray, Bedrock, vLLM for ML training and inference
• Chronon Feature store
• Kafka, postgres, redshift and snowflake for streaming and data warehousing