Machine Learning Scientist, Livehelp - Toronto, Canada - Wayfair

Wayfair

Verified Company

Toronto, Canada

1 week ago

Posted by:

Sophia Lee

beBee Recruiter

Description

This is a hybrid role located at our downtown Toronto office.

Wayfair Machine Learning Science powers automation and decision support across all Wayfair business units. Our algorithms tackle a varied and broad spectrum of challenges in the Wayfair marketplace; from empowering suppliers to easily add products to our catalog, to enabling our customers to discover and purchase a vast and diverse assortment of home goods.

ML LiveHelp builds and maintains products supporting Wayfair's live customer contacts and customer order issue remediation which outperform traditional software by utilizing adaptive, data-driven technologies. We have been a partner in the success of our business by saving millions of dollars each year through infusing machine learning into live customer interactions with Wayfair's sales and customer service teams. We use our intelligent customer intent prediction systems to automate contacts through a virtual assistant, surface relevant information to agents during a live contact, and ensure that our customers reach the best available agent to meet their needs. We strive to go beyond by exploring new opportunities to use machine learning across the Sales and Service organizations to better help over 100 million customers at Wayfair.

The Machine Learning Scientist role within the LiveHelp ML team at Wayfair will develop and deploy machine learning models that power algorithmic decision-making improving the live interactions between our customers and Wayfair's sales and service teams.

To do so you will work with a great team of ML Scientists, business stakeholders and ML Engineers to tackle problems like:

Identifying the best way to leverage multilingual language models (like XLM-R) to expand access to our virtual assistant in a scalable way.
Reducing the size of language models (like distilBERT) to maintain high accuracy with fast inference times.
Leveraging feature importances from our CatBoost and XGBoost models to understand model behavior and identify useful new feature categories.
Optimizing decision making across multiple predictive models (customer intent, expected cost/revenue) to improve the routing of our contacts.

We work as a team to deploy robust real time and batch models into production, and measure their impact using A/B testing and causal inference techniques.

This helps us ensure that we are continuously improving customer satisfaction by optimizing our contacts to be more personalized and cost effective.

What You'll Do

Own the data science life cycle from scoping to prototyping, testing, deploying, measuring value and iterating.
Partner closely with various business and engineering teams to drive the integration of our model outputs and algorithmic decisionmaking systems to integrate ML products into technical platforms and deploy realtime models and services to serve millions of customers.
Extend existing ML libraries and frameworks for scalable model training and deployment.
Partner with Production and Analytics teams to help with designing A/B testing experiments and guide business decisions using model outputs and findings.

What You'll Need

3+ years of experience in a quantitative or technical work environment, and advanced degree (MS, PhD) in a quantitative field (e.g. mathematics, economics, computer science, engineering, physics, neuroscience, operations research, etc)
Thorough command of general data science and machine learning techniques
Proficient at Python (knowledge of other programming languages like R, Java, etc. is a plus)
Comfortable with SQL and ability to wrangle data from various sources (the experience of using cloud based tools would be a plus)
Machine Learning experience (such as supervised/unsupervised learning, deep learning, Reinforcement Learning, etc.)
A background in operations research or experience with simulation and optimization is preferred
Ability to thrive in a dynamic environment where there can be degrees of ambiguity
Demonstrable commercial experience in productionenvironmentdriven ML design (e.g. having deployed at least one ML solution to scale, etc.)
Ability to work on crossfunctional projects and manage multiple stakeholders with competing priorities
Good understanding of experimental techniques for the design of A/B tests to measure the impact of initiatives
Communication skills that can influence across organizations and levels

About Wayfair Inc.
Wayfair is one of the world's largest online destinations for the home.

Whether you work in our global headquarters in Boston or Berlin, or in our warehouses or offices throughout the world, we're reinventing the way people shop for their homes.

Through our commitment to industry-leading technology and creative problem-solving, we are confident that Wayfair will be home to the most rewarding work of your career.

If you're looking for rapid growth, constant learning, and dynamic challenges, then you