In our recent TWIMLcon conversation with Solmaz Shahalizadeh, VP of Commerce Intelligence at Shopify, she shared her journey from deploying Shopify’s first machine learning model four years ago to now having over 100 models in production across all aspects of the business.
Though she had much to share, one quote that stood out was, “If you’re serious about your data, invest in your platform.”
To ensure investment in the right technologies, people, and processes, Solmaz and her team first set out to define a set best practices through a bit of trial and error. After the development and deployment of several models, she and her team were able to identify the best practices that mattered most:
- Focus machine learning on solving impactful business problems
- Seamlessly integrate machine learning into the general product development experience
- Invest in how you train, deploy, and monitor models and features
- Monitor the impact of models on users and products by shadowing them in production before going live
The team at Shopify then invested in an ML platform that unified these best practices. They also further invested in dimensional modeling to understand the business and developed feature stores and data discovery methods aimed at ensuring a shared understanding of the business across the company.
Today, Shopify has reached a level of scale and shared understanding where they can quickly pivot as the world, critical features, and data distributions change. By leveraging both systems and humans that are aware of changes in the broader business context and how these changes impact the company’s ML models, the company has been able to successfully navigate the challenges of the ongoing covid-19 pandemic.
To hear more about Solmaz’s perspective on how to lead and organize highly effective and scalable data science teams and her thoughts on being nimble and innovative in a rapidly changing world check out TWIMLcon On Demand. By revisiting the great content from this year’s TWIMLcon, you will learn from world-class presenters and panelists from teams leading the application of AI and Machine Learning at companies like Netflix, Toyota, LinkedIn, Spotify, Google, Walmart, iRobot, Adobe, Intuit, Yelp, Salesforce, Prosus Group, Palo Alto Networks, Microsoft, Qualcomm, and more. TWIMLcon On Demand’s 20+ hours of presentations, workshops, and discussions will provide you with a practical blueprint for delivering machine learning efficiently and at scale. To explore this great content and learn more about building smarter, innovating faster, and avoiding costly mistakes across end-to-end ML model production, visit twimlcon.com/ondemand.