Inbal Budowski-Tal & Dana Racah

Customer Experience Through Machine Learning Metrics

EverC

Inbal Budowski-Tal & Dana Racah

Customer Experience Through Machine Learning Metrics

EverC

Bio

Inbal is the Director of AI at EverC – a FinTech startup in the world of Risk Analysis. She is a versatile researcher: be it the field of risk intelligence at EverC, biological data at a pharma startup, semi-structured user data at Microsoft, or 3D protein structures for her Ph.D. Her passion for data is agnostic to its type, and she believes that with a basic understanding of how to represent data, model it, and measure it, it would become applicable in various business domains.

Dana is a Senior Data Scientist at EverC. Before joining EverC, she was a Data Scientist at Natural Intelligence and was responsible for building models that assisted in the understanding and optimization of user experience. She is a strong believer in the end-to-end approach: Data Scientists participate in defining the business problem and acceptance criteria through developing and testing models, and develop the web-services that serve our models in production. She holds an MSc in Computer Science and a BSc in Computer Science and Computational Biology from the Hebrew University.

Bio

Inbal is the Director of AI at EverC – a FinTech startup in the world of Risk Analysis. She is a versatile researcher: be it the field of risk intelligence at EverC, biological data at a pharma startup, semi-structured user data at Microsoft, or 3D protein structures for her Ph.D. Her passion for data is agnostic to its type, and she believes that with a basic understanding of how to represent data, model it, and measure it, it would become applicable in various business domains.

Dana is a Senior Data Scientist at EverC. Before joining EverC, she was a Data Scientist at Natural Intelligence and was responsible for building models that assisted in the understanding and optimization of user experience. She is a strong believer in the end-to-end approach: Data Scientists participate in defining the business problem and acceptance criteria through developing and testing models, and develop the web-services that serve our models in production. She holds an MSc in Computer Science and a BSc in Computer Science and Computational Biology from the Hebrew University.

Abstract

A segment of the Data Science team’s work is to continuously update and improve production models. In one of EverC’s projects, we produced a new model that beats the existing model in both precision and recall! However, once this spectacular, innovative model was deployed we started receiving complaints from customers, claiming their experience was degraded. What went wrong?

In this roundtable, we will focus on the gap between traditional machine learning metrics and the customers’ experience and suggest ways to bridge over that gap.

Abstract

A segment of the Data Science team’s work is to continuously update and improve production models. In one of EverC’s projects, we produced a new model that beats the existing model in both precision and recall! However, once this spectacular, innovative model was deployed we started receiving complaints from customers, claiming their experience was degraded. What went wrong?

In this roundtable, we will focus on the gap between traditional machine learning metrics and the customers’ experience and suggest ways to bridge over that gap.

Discussion Points

  • Data scientist role definitions – full stack data scientists vs. specialisations
  • Pure data science teams vs embedded teams
  • Data science reporting lines
  • Professional and personal development in embedded teams

Discussion Points

  • Data scientist role definitions – full stack data scientists vs. specialisations
  • Pure data science teams vs embedded teams
  • Data science reporting lines
  • Professional and personal development in embedded teams