Building a Feature Store to Reduce the Time to Production of ML Models

apply(conf) - Apr '21 - 10 minutes

Within the world of software development and especially in the area of Machine Learning, times assigned to analysis, development and deployment are key to the success of an organization. Experts’ work in the Feature Engineering process is one of the tasks that takes the longest when building a Machine Learning model and in which saving time becomes a differentiating factor. In this context, having a Feature Store is crucial for any organization that wants to improve its “time to market”. At Mercado Libre we are not exempt from this problem, that is why in this lightning session I will approach a solution proposal of Feature Store as well as the benefits that we have gained by implementing it in our organization.

Emiliano Martin

Sr. Software Engineer

Mercado Libre

My name is Emiliano Martín. I live in Villa María, Córdoba, Argentina. I am a Software Engineer graduated from Universidad Tecnológica Nacional of Villa María. I started my professional career in 2015 at Mercado Libre as a BackEnd Java developer in the Fraud Prevention area of MercadoPago. In 2017 I moved to the Anti-Money Laundering area where apart from BackEnd I also worked as a FrontEnd developer with NodeJS and ReactJS.

From October 2019 to the present time I have worked as MLOps in the Machine Learning Technology area.
I am married and have a daughter. In my free time I love spending time with my family and playing video games. My favorite sports are football and basketball. I am passionate about Software Development