You are viewing content from a past/completed QCon -

Presentation: Zero to Production in Five Months @ ThirdLove

Track: Modern CS in the Real World

Location: Pacific LMNO

Duration: 1:40pm - 2:30pm

Day of week:

Slides: Download Slides

Level: Intermediate - Advanced

Persona: CTO/CIO/Leadership, Data Engineering, Data Scientist, Developer

This presentation is now available to view on

Watch video with transcript


At ThirdLove, we believe that every woman should be comfortable and confident every day. For half of the world’s population that means a bra whose straps stay put, their cup mold to their shape and is the perfect style for the occasion. In this talk, we will discuss how we built our first machine learning recommendation algorithm that predicts bra size and style. We will cover two broad topics - the first being the challenge of working with real-world data where there is no truth flag. Then we will talk about the tradeoffs associated with key decisions we made around design, implementation and testing.

Speaker: Megan Cartwright

Head of Data Science @ThirdLove

Megan is passionate about discovering causal relationships closely followed by a love of warp drive. Now she’s enjoying the challenge of predicting the best bra for any woman. She entered data science via physics where she developed algorithms to predict energy transfer across the solar system. She’s from the PNW, but loves the CA sun.  She earned her PhD in space plasma physics from UCLA and has worked in e-commerce at One Kings Lane and Eventbrite.

Find Megan Cartwright at