9 mustn’t-miss machine learning sessions at Next ‘19
From predicting appliance usage from raw power readings, to medical imaging, machine learning has made a profound impact on many industries. Our AI and machine learning sessions are amongst our most popular each year at Next, and this year we’re offering more than 30 on topics ranging from building a better customer service chatbot to automated visual inspection for manufacturing.
If you’re joining us at Next, here are nine AI and machine learning sessions you won’t want to miss.
1. Automating Visual Inspections in Energy and Manufacturing with AI
In this session, you can learn from two global companies that are aggressively shaping practical business solutions using machine vision. AES is a global power company that strives to build a future that runs on greener energy. To serve this mission, they are rigorously scaling the use of drones in their wind farm operations with Google’s AutoML Vision to automatically identify defects and improve the speed and reliability of inspections. Our second presenter joins us from LG CNS, a global subsidiary of LG Corporation and Korea’s largest IT service provider. LG’s Smart Factory initiative is building an autonomous factory to maximize productivity, quality, cost, and delivery. By using AutoML Vision on edge devices, they are detecting defects in various products during the manufacturing process with their visual inspection solution.
2. Building Game AI for Better User Experiences
Learn how DeNA, a mobile game studio, is integrating AI into its next-generation mobile games. This session will focus on how DeNA built its popular mobile game Gyakuten Othellonia on Google Cloud Platform (GCP) and how they’ve integrated AI-based assistance. DeNA will share how they designed, trained, and optimized models, and then explain how they built a scalable and robust backend system with Cloud ML Engine.
3. Cloud AI: Use Case Driven Technology (Spotlight)
More than ever, today’s enterprises are relying on AI to reach their customers more effectively, deliver the experiences they expect, increase efficiency and drive growth across their organizations. Join Andrew Moore and Rajen Sheth in a session with three of Google Cloud’s leading AI innovators, Unilever, Blackrock, and FOX Sports Australia, as they discuss how GCP and Cloud AI services, like the Vision API, Video Intelligence API, and Cloud Natural Language have made their products more intelligent, and how they can do the same for yours.
4. Fast and Lean Data Science With TPUs
Google’s Tensor Processing Units (TPUs) are revolutionizing the way data scientists work. Week-long training times are a thing of the past, and you can now train many models in minutes, right in a notebook. Agility and fast iterations are bringing neural networks into regular software development cycles and many developers are ramping up on machine learning. Machine learning expert Martin Görner will introduce TPUs, then dive deep into their microarchitecture secrets. He will also show you how to use them in your day-to-day projects to iterate faster. In fact, Martin will not just demo but train most of the models presented in this session on stage in real time, on TPUs.
5. Serverless and Open-Source Machine Learning at Sling Media
This session covers Sling’s incremental adoption strategy of Google Cloud’s serverless machine learning platforms that enable data scientists and engineers to build business-relevant models quickly. Sling will explain how they use deep learning techniques to better predict customer churn, develop a traditional pipeline to serve the model, and enhance the pipeline to be both serverless and scalable. Sling will share best practices and lessons learned deploying Beam, tf.transform, and TensorFlow on Cloud Dataflow and Cloud ML Engine.
6. Understanding the Earth: ML With Kubeflow Pipelines
Petabytes of satellite imagery contain valuable indicators of scientific and economic activity around the globe. In order to turn its geospatial data into conclusions, Descartes Labs has built a data processing and modeling platform for which all components run on Google Cloud. Descartes leverages tools including Kubeflow Pipelines as part of their model-building process to enable efficient experimentation, orchestrate complicated workflows, maximize repeatability and reuse, and deploy at scale. This session will explain how you can implement machine learning workflows in Kubeflow Pipelines, and cover some successes and challenges of using these tools in practice.
7. Virtual Assistants: Demystify and Deploy
In this session, you’ll learn how Discover built a customer service solution around Dialogflow. Discover’s data science team will explain how to execute on your customer service strategy, and how you can best configure your agent’s Dialogflow “model” before you deploy it to production.
8. Reinventing Retail with AI
Today’s retailers must have a deep understanding of each of their customers to earn and maintain their loyalty. In this session, Nordstrom and Disney explain how they’ve used AI to create engaging and highly personalized customer experiences. In addition, Google partner Pitney Bowes will discuss how they’re predicting credit card fraud for luxury retail brands. This session will discuss new Google products for the retail industry, as well as how they fit into a broader data-driven strategy for retailers.
9. GPU Infrastructure on GCP for ML and HPC Workloads
ML researchers want a GPU infrastructure they can get started with quickly, run consistently in production, and dynamically scale as needed. Learn about GCP’s various GPU offerings and features often used with ML. From there, we will discuss real-world customer story of how they manage their GPU compute infrastructure on GCP. We’ll cover the new NVIDIA Tesla T4 and V100 GPU, Deep Learning VM Image for quickly getting started, preemptible GPUs for low cost, GPU integration with Kubernetes Engine (GKE), and more.