AI and Machine Learning
Introduce efficiencies to improve operations and add intelligence for digital transformation of products or services.
- 236 Topics
- 279 Replies
I writing here about Google Cloud's work over the past eight years to bring new levels of performance, scale, reliability, and power efficiency to Google's Jupiter datacenter network.The combination of optical circuit switching and Software Defined Networking allowed google cloud to bring flexible topologies and real-time traffic engineering to the datacenter, enabling seamless upgrades and adaptation to application and network dynamics.In an industry first, Google cloud is able to develop and deploy at-scale optical circuit switching to configure network topologies through software control. Through a building-scale, all-optical backbone, Google Cloud can now support heterogeneous technology and evolving mesh topologies, orchestrating the equivalent of thousands of logical fiber moves with no application impact and no physical presence.Jupiter serves as a key enabler for Google Cloud and dozens of planetary-scale Google services. Now google cloud is able to bring this next level of ca
Today, I’ll introuduce with Google cloud AI latest libray Rax, a library for LTR in the JAX ecosystem. Rax brings decades of LTR research to the JAX ecosystem, making it possible to apply JAX to a variety of ranking problems and combine ranking techniques with recent advances in deep learning built upon JAX (e.g., T5X). Rax provides state-of-the-art ranking losses, a number of standard ranking metrics, and a set of function transformations to enable ranking metric optimization. All this functionality is provided with a well-documented and easy to use API that will look and feel familiar to JAX users. Please check out our paper for more technical details. Click on the below link to read more details.https://ai.googleblog.com/2022/08/rax-composable-learning-to-rank-using.html
Calling all storage architects and enthusiasts. Learn how to make your storage infrastructure work harder for your business. Join our digital event to discover what’s new in storage from Google Cloud. Hear from technical experts across our Storage portfolio, see new products in action with live demos, listen to how our technology is helping customers grow, and take away new ideas that will help you take advantage of Google Cloud’s latest storage innovations. Date and time : 8 September 2022 17:00- 18:15 BST Click on the following link to join the event: https://cloudonair.withgoogle.com/events/storage-spotlight
Today I will discuss how to build an ML-based network anomaly detection application for telecom networks to identify cyber security threats by using Dataflow, BigQuery ML, and Cloud Data Loss Prevention and used to help identify cybersecurity threats.In this tutorial, I'm trying to focus on google cloud documentation key point to develop an anomaly detection application intended for data engineers and scientists. Also, I assume that you have basic knowledge of the following:Dataflow pipelines built using the Apache Beam Java SDK Standard SQL Basic shell scripting K-means clusteringArchitecture Overview The following diagram shows the components used to build an ML-based network anomaly detection system. Pub/Sub and Cloud Storage serve as data sources. Dataflow aggregates and extracts features from this data tokenized with DLP API. BigQuery ML creates a k-means clustering model from these features and Dataflow identifies the outliers. ObjectivesCreate a Pub/Sub topic and a subscription
Kaluza is a UK-based technology company that provides energy retailers with real-time billing, smart grid services, and seamless customer experiences. In this blog, Tom Mallett, Sustainability Manager, Kaluza, explains how Kaluza leverages Google Cloud to improve energy visibility throughout the company. He also explores how better emissions data informs sustainability solutions that make the world’s energy greener, smarter, and more reliable.To read more information, click the link below : https://cloud.google.com/blog/topics/sustainability/kaluza-uses-google-cloud-to-help-people-reduce-emissions
Today, most robots are found in industrial settings and have been rigorously programmed to do particular jobs. What if there was a more effective way to interact with intelligent robots so they could assist us? Together with helper robots that can handle challenging and abstract tasks like "cleaning up a spilt drink," researchers and engineers from Google Research and Everyday Robots are developing the best machine learning language models.Recently, Brian Ichter and Karol Hausman, Research Scientists, Google Research, Brain Team wrote an article based on a better way to communicate with learning robots. Please click on the following link to read more details. https://ai.googleblog.com/2022/08/towards-helpful-robots-grounding.html
Tired of tuning your parameter-heavy transformers with Hyperparameter tuning algorithms (even those as advanced as Google's Vizier)? Instead, why not use transformers to tune your Hyper Parameters?Click on the following link to read more details. https://ai.googleblog.com/2022/08/optformer-towards-universal.html
who is newly practicing as a google cloud solution architect or trying to build a good banking app it might help you to understand how can use google cloud resources to build your app in the cloud easier. In this video, Developer Advocate Priyanka discusses an approach to building a modern banking platform in the cloud. Watch to learn about the power of Google Cloud and how you can also leverage the cloud to build a modern banking platform. The video contains the following topics:0:00 - Intro0:47 - The current state of banking1:35 - Banking as a service model architecture2:24 - APIs for banking3:21 - Let us know in the comments3:35 - High level overview6:06 - The architecture6:36 - Channel services7:01 - The customer applications7:13 - The integration layer (API connectors)8:00 - Event / Transformation hub8:14 - Data for enterprise use8:47 - Back end services9:16 - Core application services9:49 - Core security and operational services10:19 - Sample application code10:32 - Wrap up Extra
Develop an AI strategy for your data. Solve business-critical problems with technology for any level of ML expertise. IDC found that only 36% of their surveyed enterprises succeeded in putting completed models into production. Nearly half of the remainder hadn’t even made it past proof of concept. We’ve learned from Google’s years of experience in AI development how to overcome common challenges. With successful AI strategies, we’re seeing customers from many industries and varying levels of ML expertise enhance and solve business-critical problems with AI-driven decision making. Event Date and Time: 18 August 2022 17:00- 17:30 BST For more information and to register for the event, please click here.https://cloudonair.withgoogle.com/events/data-to-ai-webinar
The use of matrix factorizations in recommendation systems is widely. If you want to quickly and easily develop a solution to provide excellent recommendations to your clients, then you will get a good and simple starting point from this video: The above video contains following content : 00:00 Introduction00:49 Pipeline01:28 Dependencies02:20 Data02:55 Why reservations04:10 Create slots08:31 Delete slots09:00 Train model10:40 Deployment12:24 Serving12:47 Running Pipeline14:40 Deployment16:04 Recommendations Extra supporting credit: Check out the two Google articles if you're new to matrix factorization and BigQuery ML.Recommendations with implicit feedback https://cloud.google.com/bigquery-ml/docs/bigqueryml-mf-implicit-tutorial Recommendations with explicit feedbackhttps://cloud.google.com/bigquery-ml/docs/bigqueryml-mf-explicit-tutorial Note:The data used in the video demonstration and the Google BigQuery ML examples are identical. No need to reinvent the wheel in this case.
Doesn't real-time supply chain visibility sound great? To provide customers with complete data for safe, secure, and on-time freight, Cargo Signal is working with Google Cloud and Confluent.Google cloud help confluent to removes obstacles from the IoT data capture and analysis pipeline . Also Cuts potential data integration issues in half through managed data streaming services on Confluent Cloud and maximizes speed-to-market and accelerates issue resolution through flexible development infrastructure in GKE.Find out more, click on the following link:https://cloud.google.com/customers/cargo-signal
Does your organization store data in multiple places? Are you looking for ways to break down data silos and enable analytics? Are you looking to build a multi-cloud data lake on open file formats?Introducing BigLake, a storage engine that allows organizations to unify data warehouses and lakes. Watch along and learn how to perform uniform fine-grained access control and accelerate query performance across multi-cloud storage and open formats all while maintaining a single copy of your data.Take two minute to watch the following video about Big Lake
Wondering how restaurants and retail establishments use quality control or estimate demand? In this instalment of Build With Google Cloud, Priyanka Vergadia demonstrates how to create an edge solution for a hypothetical hamburger chain that wants to deploy and use uniform technology across all of its locations. Watch this demonstration to see how to update a restaurant and predict demand with Google Cloud! Chapters:0:00 - Intro1:16 - Edge Patterns - IoT1:37 - Rundown of scenarios4:13 - Anthos for Edge5:04 - Demo - Demand Forecasting8:50 - Demo - Dining Room Cleanliness10:39 - Wrap up
A managed ML training service can help you automate experimentation at scale or retain models for a production application. In this following video Prototype to Production, Developer Advocate, Nikita Namjoshi, walks through the steps required to train custom models on Vertex AI. Watch along and learn about the benefits of a managed training service that helps keep your results fresh. Chapters:0:00 - Intro0:22 - Why do I need a machine learning training service?1:26 - What are containers?2:19 - Update custom training code3:23 - Cloud storage for machine learning4:50 - Containerizing code for machine learning5:39 - Dockerfile syntax6:42 - How to store container images in Google Cloud7:21 - How to launch a training job on Vertex AI8:12 - Wrap up
In this discussion, I will focus on the reference architecture for technical decision makers who want to connect devices and build Cloud IoT apps on Google Cloud using Intelligent Products Essentials.At this moment, Manufacturers want to continuously improve their products by adding intelligent capabilities that delight customers and monetize new features. In this article, I will describe how google helps architects tasked with designing intelligent product systems on Google Cloud that are scalable, reliable, secure, and cost-effective.Google Cloud provides capabilities for connecting, ingesting, storing, analyzing, and retrieving data from products to build in artificial intelligence and machine learning capabilities, such as a personalized product ownership app or a digital twin simulation. Here I will present an overview of architecture, its high-level components, integration topics, and general design considerations. The architecture includes the following:The architecture includes
MLOps, like DevOps, is used to enhance the quality and minimise the time to market of machine learning engineering. It may improve team cooperation, increase the reliability and scalability of ML systems, and minimise development cycle times.It can keep the ML system from failing due to a failure to respond to environmental changes. Here is a detailed look at the finest cloud documentation-based MLOPs implementational architecture.This article demonstrates, using real-world examples, how you can practise MLOps and acquire ideas for constructing valuable solution infrastructure.this document covers the following topics:Understanding CI/CD and automation in ML. Designing an integrated ML pipeline with TFX. Orchestrating and automating the ML pipeline using Kubeflow Pipelines. Setting up a CI/CD system for the ML pipeline using Cloud Build. Please click on the following link to read more details:https://cloud.google.com/architecture/architecture-for-mlops-using-tfx-kubeflow-pipelines-and-
Max Saltonstall and new presenter Anu Srivastava are in the studio to discuss Vertex Explainable AI with guests Irina Sigler and Ivan Nardini. Vertex Explainable AI arose from the desire for developers to better understand how their algorithms select classifications.This categorization understanding is critical for two reasons: trusting the operation of models for business decision making and faster debugging. Also, They dicuss in GCP Podcast, why explainable models are so crucial and how Vertex Explainable AI may assist.To hear and read the transcript, click on the following podcast link:https://www.gcppodcast.com/post/episode-314-vertex-explainable-ai-with-irina-sigler-and-ivan-nardini/
If you're like many people new to Google Cloud, you've worked with other public clouds in the past.Join this event on August 24th for a friendly, factual, non-competitive exploration about what's different in Google Cloud - intended to help people more familiar with AWS and Azure overcome initial hurdles and accelerate their cloud journey.You’ll also have an opportunity to ask questions and receive answers live.
Data de-identification, a type of dynamic data masking, refers to severing the connection between the data and the person with whom it was first linked. In essence, this calls for the removal or transformation of personal identification.Data has evolved into a vital and important reservoir of value for businesses. As a result, businesses must safeguard and manage data in a secure and effective manner. Google provides Cloud Data Loss Prevention (CDLP). You may examine, check out, and de-identify your data with the aid of this completely managed service. Using CDLP, you may examine millions of pieces of data and determine which ones are sensitive and have to be de-identified or encrypted.Additionally, you may set it to automatically allow it to disguise your private information (this operation is known as de-identifying). More than a hundred infoTypes are available in CDLP. A type called an infoType is used to represent sensitive data, such as a date, a name, an email address, or a phon
How SkyTruth use Google cloud and ML help to scan the world’s oceans to detect the signs of oil slicks?
In the ocean, how many oil slicks are there? They have no idea where they are or how they got there.You want to know about how a nonprofit environmental watchdog Google Maps, Machine Learning, and Cloud Computing are being used by SkyTruth to analyse radar satellite photos in order to look for indicators of oil slicks.The conservation technology organisation SkyTruth is redefining how they track environmental impacts using Google Maps Platform, machine learning, and cloud computing, according to Mitchelle De Leon, Director of Impact and Strategic Partnerships at SkyTruth.Additionally, he provides a detailed look at SkyTruth's Cerulean, a platform that use artificial intelligence to examine hundreds of radar satellite photos each day in search of indicators of oil slicks.Cerulean will detect oil slicks from ships, offshore oil rigs, and other sources as SkyTruth intensifies the automation process. This will enable the creation of a worldwide map of oil pollution and the identification o
Without interruption, DevOps is implemented in the Kubernetes engine throughout the creation of AI/ML applications. To manage application deployment situations like "Continuous Deployment," "Blue-Green Deployments," "Canary Deployments," and more, Dev Ops procedures will frequently require numerous deployments. In this article, I'll go over some fundamental container scaling and management techniques so you can complete these frequent tasks when using several heterogeneous deployments.What I will disscus: Kubernet engine kubectl tool command Create and manage deployment yaml files How can, update, and scale deployments Updating deployments and deployment stylesI hope you have some basic knowledage about Kubernet engine and DevOps theory. About Heterogeneous deployments Heterogeneous deployments usually involve the connection of two or more distinct infrastructure environments or regions to respond to a specific technical or operational need. Heterogeneous deployments are known as "hybr
Whether it’s a product launch, new store opening, limited-time sale, or holiday, peak traffic events are critical for your business. Don’t just hope that they’ll run smoothly! Join this event for Ask Anything session on August 25th to learn how to successfully plan for peak traffic and launch events to avoid any disruptions for your customers and your business.The experts will cover the three key components in preparing for a successful event:Preparation: Activities that can help you prepare for your event include an architecture review, load testing, quota limits and/or capacity planning - just to name a few. Execution: As your event begins you’ll need to closely monitor and react accordingly. Analysis: After your event is completed, analyze what went well, what didn’t, and how to improve for future events.As always, you’ll also have an opportunity to ask the experts questions and receive answers live. With this series of Ask Me Anything events, it's gogole cloud goal to provide a tru
Join on August 17th at 9AM PT for a live session with Google Cloud experts on Dataflow observability, monitoring, and troubleshooting.Google Cloud expert will provide an overview of all the observability experiences in Dataflow and discuss more on the new features. They will also look at the common symptoms or issues reported by customers for streaming pipelines and discuss how to use the new observability features and tools to troubleshoot those.To participate in the live session on August 17 at 9 a.m. Pacific Time, click on the video link below. Video chapter:03:02 Evolution of data and the value gap06:21 Why Dataflow?08:44 What is Dataflow?10:25 What is observability? Core Dataflow observability features:11:48 Job visualizers14:38 Job metrics21:06 Cloud Error Reporting integration22:19 Cloud Profiler integration23:33 Dataflow insights (recommendations)25:21 Datadog integration26:54 Where to start?29:23 Live demo Troubleshooting common scenarios48:31 Job slows down with increase in
Are you a startup striving for growth and innovation? Are you looking to innovate, revamp and make the best use of your cloud infrastructure? Or are you simply looking to unbox, learn & grow your business with a leading cloud services provider?We at Google Cloud are here to help you learn & understand GCP cloud services and use them to improve your technical landscape. Our focus is to empower growing startups like yours with the right training, tools, technologies, interactive product workshops and support.We are thrilled to announce 60 minutes of monthly interactive sessions on GCP Products & Services conducted by Google Customer Engineers.These detailed sessions & workshops have been especially curated for your organization’s needs and cater to specific use cases that you are working on followed by Q&A discussions with Google Cloud experts.Our customer engineers will be available for 1:1 engagements with startups wanting to discuss architectures, migrations, solut
Recently Google cloud worked with Apollo 24|7, the largest multi-channel digital healthcare platform in India, to build the key blocks of their CDSS solution. It is a crucial piece of healthcare technology, the Clinical Decision Support System (CDSS), that analyses data to assist healthcare practitioners in making decisions regarding patient care.Google cloud AI helped them to parse the discharge summaries and prescriptions to extract the medical entities. These entities can then be used to build a recommendation engine that would help doctors with the “Next Best Action” recommendation for medicines, lab tests, etc.Vertex AI, a Python package comprising spaCy models for analysing biomedical, scientific, or clinical literature, is used by Google cloud to conduct research based on ScispaCy.To learn more about Google Cloud and the Apollo 24|7 Building Clinical Decision Support System (CDSS), click the link below:https://cloud.google.com/blog/products/ai-machine-learning/apollo-24-7s-cdss-
Login to the community
Social LoginLogin With Your C2C Credentials
Enter your username or e-mail address. We'll send you an e-mail with instructions to reset your password.