IWDS 25: Building and deploying reproducible ML pipelines by Soledad Galli

#########
To avoid disappointment, please remember to register at: https://skillsmatter.com/meetups/12886-building-and-deploying-reproducible-ml-pipelines
#########

Hi everyone,

We are having an exciting September full of ML opportunities for you.

In this talk, Soledad Galli will tell us about her journey into Data Science and discuss the steps and challenges involved in putting a machine learning model into production. She will describe how we can set up an effective and reproducible machine learning pipeline, highlight the challenges to obtaining reproducible models.

About Soledad Galli:

Soledad is a Lead Data Scientist with experience in Finance and Insurance, and a Udemy instructor of machine learning courses. Soledad has 3+ years of experience in data science and analytics in finance and insurance, and 10+ years of experience in scientific research in academia. While working in finance and insurance, Sole Researched, developed and put into production machine learning models for insurance motor Claims, to assess Credit Risk and to prevent Fraud.
Having transitioned from academia to data science, Soledad is passionate about enabling data scientists and academics to transition into the field, and helping data scientists increase their breadth of knowledge. Over the last 2 years, Soledad shared insights in blogs and talks for the data science community. She also created 3 online courses on machine learning which are live in Udemy, enrolled 8000+ students worldwide and continue to receive excellent reviews from students. Soledad is passionate about extracting meaningful information from data and supporting people and organisations make solid and reliable data driven decisions.

Motivation of this talk:

Deployment of machine learning models, or simply, putting models into production, means making our models available to other systems of our organisation. By deploying models, other systems can send data to them and get their predictions, which are in turn populated back into the organisation software and applications. Through machine learning model deployment, we can begin to take full advantage of the models we built.

When we think about data science, we think about how to build machine learning models, about which algorithm is more predictive, how to engineer the variables or which variables we should use to make the models more accurate. But, how to actually consume those models is often overlooked at the time of design. And yet this is the most important step in the machine learning pipeline.