Você está visualizando o site de uma EDIÇÃO PASSADA do QCon.

Palestra: Metrics Driven Machine Learning Development at Salesforce Einstein

Track: Armazenamento e Processamento de Big Data

Sala: 4 São Francisco

Horário: 5:20pm - 6:05pm

Dia da semana: Segunda-feira

Slides: Download Slides

Nível: Intermediário

Persona: Cientista de Dados, Desenvolvedor(a) Programador(a), Desenvolvedor(a) Sênior, Líder Técnico(a), Product Owner

Apresentação em Inglês

Share this on:

This presentation is now available to view on InfoQ.com

Assista a palestra com transcrição

Pontos Principais

  • The AutoML metrics framework enables the Einstein team to monitor and troubleshoot thousands of models running and production;
  • Metrics can automatically alert us of any modeling issues across the platform;
  • They are an invaluable part of the development process- by tracking metrics in modeling experiments we can test and evaluate new features before deploying them to production.


The Einstein Prediction Builder modeling pipeline automates all steps of the end-to-end modeling process, from data auditing and feature engineering to model selection, for thousands of models. Although powerful, automated machine learning pipelines are inherently a black box that can be notoriously difficult to troubleshoot. This talk walks through a Data Science perspective of using monitoring and alerting, well-established practices in traditional engineering, to develop our modeling pipeline. We will discuss how we track data and modeling metrics at each stage in the pipeline to identify data and modeling issues and to raise alerts for issues affecting models running production. Furthermore, we will cover how this metrics framework is instrumental in helping to develop new features in a data-driven manner.

Palestrante: Eric Wayman

Senior Data Scientist at Salesforce

Eric Wayman is a Senior Data Scientist at Salesforce. As a member of the Einstein AI platform team, he works on developing the automated machine learning Pipeline for the recently released Einstein Prediction Builder, which helps Salesforce administrators leverage their Salesforce data to make predictions for use cases tailored to their individual needs. Before joining Salesforce, Eric worked as a Data Science Consultant at Pivotal Software and also did research in Probability and Stochastic Processes at UC Berkeley where he received his Ph.D. in Mathematics.

Find Eric Wayman at

Tracks 2019