Eric Siegel | The AI Playbook | Talks at Google

TL;DR

Eric Siegel presents a comprehensive six-step playbook for running enterprise machine learning projects successfully through to deployment.

Transcript

[MUSIC PLAYING] BRETT DURRETT: Welcome to Talks at Google. I'm Brett Durrett. And today, we're welcoming Eric Siegel to Talks at Google. Eric has been in machine learning for over 30 years. He's the founder and CEO of Gooder AI, as well as the founder of the long-running Machine Learning Week, a conference that's coming up in two weeks in Phoenix. ... Read More

Key Insights

• 😌 Machine learning enables businesses to make accurate predictions, but its value lies in actual deployment and integration into existing operations.
• 🖤 The lack of a well-adopted framework and deep collaboration is the main reason why most machine learning projects fail to deploy.
• 🥅 There are numerous untapped opportunities for machine learning across industries, but each project should be tailored to the specific needs and goals of the organization.

Questions & Answers

Q: Why do most machine learning projects fail to achieve deployment?

Most projects fail due to a lack of a well-adopted framework and a disconnect between data scientists and business stakeholders.

Q: What is the importance of deep collaboration between data scientists and business stakeholders?

Deep collaboration ensures that there is a shared understanding of the goals, metrics, and outcomes of the machine learning project, resulting in successful deployment.

Q: What are some examples of machine learning use cases within different industries?

Examples include marketing campaigns, credit scoring, fraud detection, customer churn prediction, as well as various applications in healthcare, manufacturing, and finance.

Q: What is the biggest challenge in machine learning deployment?

The biggest challenge is often the preparation of data, which must be in the right format and contain the necessary historical examples for training the model.

Q: Are there any ongoing research partnerships between educational institutions and corporations focused on improving machine learning adoption?

The focus should be on providing education and resources for non-data scientists to understand the concepts and possibilities of machine learning, as well as bridging the gap between data scientists and business stakeholders.

Summary & Key Takeaways

• Machine learning is important because it enables businesses to make accurate predictions which drive large-scale operations.

• However, most new machine learning projects fail to achieve deployment due to a lack of a well-adopted framework for running these projects.

• The BizML (Business Machine Learning) framework emphasizes the need for deep collaboration between data scientists and business stakeholders to ensure successful deployment.