Tech Lead, Data Science 805 views

The ML Engineering team within the Marketing Data Science team at Wayfair develops scalable data processing platforms and deploys hundreds of machine learning models that power algorithmic decision-making across dozens of marketing channels and customer touchpoints.

Data is at the heart of everything they do and ML data engineering is crucial to their ability to scale as they train and deploy the next generation of ML products at Wayfair that power the way millions of customers interact with them. You will be processing petabytes of un/structured first party and third party data and building scalable modeling pipelines that help them evaluate the long-term value of the millions of products they offer and of actions the business can take. The tools they’ve built thus far and the new capabilities they have on their roadmap aim to redefine how they think about label and feature generation, model development, deployment, and monitoring.

Above all, you’ll get to work on problems that are both intellectually-challenging and drive real, measurable impact, first and foremost, for their customers – and as a result for Wayfair at large. To get a better sense of the type of projects they actually work on, check out their Data Science & Machine Learning blog posts here!

What You’ll Do

• • Build highly scalable distributed data processing platforms that evaluate the long-term incremental value of their millions of offerings and of actions that the business can take.
  • Collaborate with other data scientists to build high-quality ML models that can robustly scale up to large volumes in production.
  • Partner closely with various business & engineering teams to drive the integration of their model outputs & algorithmic decision-making systems into existing production systems
  • You’ll be a builder of tools, software, and microservices that enhance or streamline various steps or challenges within the data science workflow & their tech stack
  • Extend existing ML libraries and frameworks for scalable model training & deployment
  • Be obsessed with the customer and maintain a customer-centric lens in how we frame, approach, and ultimately solve every problem they work on.

Example Projects

• • Building highly scalable distributed data processing platforms that evaluate the long-term incremental value of various customer actions (such as downloading their app, or signing up for their credit card, etc.)
  • Build a low dimensional representation of customers to enable better personalization & frictionless product discovery for any given customer. This problem is both highly impactful and intellectually-challenging given they offer millions of stylistically unique products on their platform!

What You’ll Need

• • 3+ years of experience working as a professional software developer, ML Engineer, or Data Scientist (with strong engineering skills & interest in software development)
  • BSc, MS, or Ph.D. in a quantitative field (e.g. mathematics, computer science, engineering, operations research, physics, economics, neuroscience, etc.)
  • Proficient in OOP programming; experience in Python, Java, etc. and exposure to some of Python’s ML ecosystem (NumPy, panda, sklear, tensorflow, etc.)
  • Proficient in large-scale data processing and distributed systems (HDFS, Spark, etc.)
  • Experience with CI/CD tools (ex Jenkins or equivalent), version control (Git), orchestration/DAGs (Airflow, Luigi, Kubeflow, or equivalent)
  • Exposure to machine-learning model lifecycle: training, evaluation, serving
  • Interest in deploying machine-learning models as scalable services
  • Desire to work in a collaborative environment focusing on continuous learning; writing blog posts, participating in tech talks, conducting code reviews, etc.
  • You don’t have to be an expert in all or any of the above areas but they need someone with a passion for learning and growing as a software developer and ML engineer

It’s a Bonus To Have

• Interest in causal inference techniques nice to have
• Experience with any of R, Docker, GCP, Kubernetes, Snowflake nice to have