The Google Cloud Applied ML Summit brought together the world’s leading professional machine learning (ML) engineers and data scientists to explore the latest cutting-edge AI tools for developing, deploying, and managing ML models at scale.

Agenda

6:30-6:40 pm Welcome

From research to ready --

At Google Cloud, we believe the products we bring to market should be strongly informed by our research efforts across Google and other Alphabet subsidiaries. For example, Vertex AI was ideated, incubated, and developed based on the pioneering research from Google’s research entities. Features like Vertex AI Neural Architecture Search (NAS) and Vertex AI Matching Engine were born out of discoveries made by Google’s researchers, were internally tested and deployed, and were shared with data scientists around the globe as an enterprise-ready solution, each within a matter of a few short years. Join this session to explore the innovation cycle that brings AI and machine learning tools from research to ready, and how data scientists and ML engineers can get their hands on what's next.

Running Time 12 minutes

End-to-end AutoML for model prep --

Explore the latest innovations born of Google Brain research and being made available to enterprises in an accelerated time frame. AutoML enables developers at all levels of machine learning expertise to train high-quality models specific to their business needs in minutes. Vertex AI Forecast generates highly accurate predictions within just two hours of training time and with no manual model tuning. In this session, learn about the latest innovations within AutoML and Vertex AI Forecast, and get insights from real-world customers who have already benefited from these solutions.

Running Time 12 minutes

Get into production faster with end-to-end MLOps --

According to analysts and customers alike, machine learning operations (MLOps) capabilities are crucial to enabling organizations to get value out of their ML investments – addressing challenges in deployment, scaling, and versioning efforts. In this session, discover how groundbreaking updates to Google's end-to-end MLOps services such as Vertex AI Pipelines, Vertex AI Model Registry, and Vertex ML Metadata enable practitioners to seamlessly perform tasks like running a continuous training pipeline, deploying a model, and monitoring predictive performance of an ML model for maximum organizational impact.

Running Time 12 minutes

Demo: Building a question answering system with Vertex AI --

In this demo, we’ll show you how to build a question answering system based on your own data using Vertex AI.

Running Time 16 minutes

You trained a model, now what? --

When it comes to AI, we see three trends: 1) experiments are easy but production is hard, 2) data scientists struggle to explain business impact resulting in a failure to secure executive support for investments, 3) adoption is limited because people don’t understand or trust AI. Join this session to hear why AI teams need to go beyond the technology and have the right organizational strategy so your hard work pays off. We’ll share examples and lessons learned from our work and how it links to Google AI and Machine Learning technologies such as Vertex AI and BigQuery ML.

Running Time 10 minutes

7:50 until closing - Open Discussion