Letgo and Clarity AI are partnering to share their knowledge, tools and best practices with Madrid’s Data Science community.

1st talk: Search & relevance at scale for online classifieds

Summary: letgo is a two-sided marketplace to buy and sell second-hand goods locally. Search relevance for local marketplaces faces challenges that are significantly different from the ones in web or e-commerce search: each listing is unique and the catalog is extremely volatile as listings get sold and posted continuously. In addition, listing descriptions, being user generated content, are usually short and noisy.

We will discuss those differences and present the solutions we developed at letgo with two main focuses: How the search platform was designed in order to scale to millions of daily users, and how relevance was improved. We will also survey techniques to improve precision and recall, and how to evaluate offline and online candidate changes in the system.

2nd talk: How to set up a Data Science team: Tools, sweat and tears.

Summary: In this talk, I'll share my experience setting up Data Science teams in Netflix, Miroculus, Airbnb and Clarity. More specifically, the lessons I've learned about which skills you should look for in a Data Scientist, how a Data Science team should fit within an organization, how to adapt the Agile Methodologies to the Data Science and which tools and best practices can help the Data Science teams to perform effortless in their activity. While we are at it, I'll also discuss some IRL anecdotes, show you my favourite tools, and confess more than one mistake...

Speaker: Antonio Molins is the Chief Data Scientist of Clarity AI, a startup focused on improving the efficiency of capital allocation in order to optimize social impact. Prior to that, he spent years in Netflix, where he went from developing recommendation algorithms to be product owner of all recommendation algorithms in the product. Antonio has also worked in Airbnb, where he optimize marketplace search; at Miroculus, where he designed predictive models for early cancer diagnostics; and at Goldman Sachs, where he developed and validated risk models for commodity derivatives. He holds a PhD in Medical and Electrical Engineering from the Harvard/MIT division for Health Sciences and Technology, a MS in Electrical Engineering and Computer Science from MIT, and a Superior Telecommunications Engineer degree from UPM.