Oracle adds generative AI capabilities to its HR solutions - Back End News

Oracle has integrated generative AI-powered capabilities into its Oracle Fusion Cloud Human Capital Management (HCM), aimed at improving HR processes and productivity. The new capabilities, supported by Oracle Cloud Infrastructure (OCI), use large language models to reduce the time required for tasks, enhance employee experience, improve accuracy of workforce insights, and increase business value. Customers can refine the generative AI models for their specific business needs using their own data. The AI capabilities streamline processes and automate content generation, allowing employees and HR leaders to author content more efficiently. Oracle Cloud HCM is a complete solution connecting all HR processes and leveraging built-in AI for workforce analysis and recommendations. #OracleHCM #AIpoweredHR #WorkforceProductivity #softcorpremium

Getting started with Google’s PaLM API | InfoWorld

Google has developed a large language model called Pathways Language Model 2 (PaLM 2), which is trained on a massive dataset of text and code. PaLM 2 can perform various tasks such as word and code completion, question answering, summarization, and sentiment analysis. It is a core component of Google's AI chatbot and AI assistant framework. The PaLM API is available within the Vertex AI platform and as a standalone API endpoint. Developers can access the API through the Vertex AI platform or the Trusted Testers program. A tutorial is provided on how to use the PaLM API in the Vertex AI platform to generate text. #GoogleAI #PaLM2 #VertexAI #softcorpremium

Mastering Azure Subscriptions Management: Best Practices for Success for Startups or Beginners

This blog post provides best practices for designing and managing Azure subscriptions. It emphasizes the importance of analyzing workloads and clients to determine the number and organization of subscriptions needed. Azure Management Groups are recommended for organizing and managing subscriptions effectively. Azure Policies and Blueprints are discussed as tools for applying governance and security controls to subscriptions. The concept of an Azure landing zone is introduced as a scalable and secure foundation for Azure workloads. The post concludes by highlighting the importance of following Azure best practices for successful subscription management. #AzureManagement #SubscriptionManagement #AzureBestPractices #softcorpremium

“Unit testing” with Azure AD B2C - The new control plane - Medium

The article discusses the use of unit testing in Azure AD B2C for custom policies. The author suggests creating a unit test Technical Policy that can be used to test the logic of the custom policy before all the required components, such as REST API and databases, are ready. This approach allows for testing different flows by changing the values of flags. The author also mentions an improved method using query strings to avoid multiple policy uploads. #AzureAD #unittesting #custompolicies #softcorpremium

6 best practices to keep Kubernetes costs under control | InfoWorld

This article discusses the challenges of managing costs in a Kubernetes infrastructure and provides tips and best practices for effective cost management. It highlights the importance of right-sizing cloud instances, using multi-tenancy, implementing cost visibility and monitoring, and adopting cost management solutions. The article also suggests exploring spot instances and optimizing resource allocation through profiling applications and using Kubernetes scaling features. It emphasizes the decision of when to use dedicated or shared clusters and the need for security and governance in a multi-tenant deployment. The article also emphasizes the importance of implementing cost optimization policies, such as deleting unused and under-utilized resources, and automating tasks to reduce costs and improve product development. It recommends using purpose-built Kubernetes cost management tools and native Kubernetes cost management solutions. Overall, following these best practices will help ensure effective cost management in Kubernetes deployments. #KubernetesCostManagement #CostOptimization #CloudInfrastructure #softcorpremium

Databricks’ $1.3 billion MosiacML acquisition to boost generative AI offerings | InfoWorld

Data analytics provider Databricks is acquiring model-training software provider MosiacML for $1.3 billion to enhance its offerings. MosiacML's models, training and inference capabilities will be added to Databricks' data lakehouse platform for enterprises to develop generative AI applications. MosiacML's models are open source and offer a \"zero human intervention\" feature, making the training systems automated. MosiacML's entire team of over 60 employees, including CEO Naveen Rao, will move to Databricks to continue developing more foundation models. The deal includes retention packages for MosiacML employees. MosiacML's existing customers will still be able to access their LLMs and inference offerings. #DatabricksAcquiresMosiacML #AItraining #DataLakehousePlatform #softcorpremium

Let’s Architect! Governance best practices | AWS Architecture Blog

The AWS Architecture Blog discusses best practices for governance in AWS environments, focusing on compliance, security, and operational efficiency. It highlights the importance of governance in enabling organizations to leverage AWS services while mitigating risks. The blog also features case studies and videos on how organizations like Global Payments use AWS services to scale their operations while maintaining governance and control. It covers topics such as AWS Service Catalog, AWS Organizations, infrastructure as code, and centralized logging on AWS. The blog concludes by mentioning upcoming discussions on DevOps best practices. #AWSGovernance #CloudCompliance #OperationalEfficiency #softcorpremium

GPT Making higher education affordable, personalised, inclusive, and reflective - Microsoft Community Hub

This blog discusses the implementation of a framework for co-pilot AI and human collaboration in industry and jobs, focusing on education and research papers. It explores the use of prompts, Bloom's taxonomy, and GPT for problem-solving and developing research papers. The goal is to make education affordable and tailored. The blog also mentions the influence of the Oxford Tutorial system in this aspiration. The author welcomes comments and feedback from educators and invites connections on LinkedIn. #AIandEducation #CollaborationInIndustry #TailoredEducation #softcorpremium

Document Understanding with Large Language Models on Google Cloud | by EnricoPicari | Google Cloud - Community | Jul, 2023 | Medium

This article discusses the use of large language models for document understanding in situations with limited or no data. It explores the performance of these models, specifically the IT5 model, in tasks such as named entity recognition and classification. The article also discusses the challenges and processes involved in document understanding pipelines. It mentions the use of Zero-Shot Learning and Few-Shot Learning approaches, as well as the potential of fine-tuning models for improved performance. The author suggests integrating the IT5 model into the Document Understanding pipeline architecture in a Google Cloud Platform project. It also mentions the availability of Google Cloud's capabilities, such as Cloud Run, for leveraging large language models in natural language understanding applications. #DocumentUnderstanding #LanguageModels #GoogleCloud #softcorpremium

Blocking API Key Access in Azure OpenAI Service | Journey Of The Geek

The author discusses how to prevent usage of API keys in Azure OpenAI Service. They suggest using Azure Active Directory authentication instead and tightly controlling API key usage. They mention the use of Azure API Management and advanced logging patterns to monitor API key usage. The author also discusses the use of Azure OpenAI Studio and the need to block API key-based authentication for certain instances. They provide instructions on how to disable API key-based authentication and mention the limitations and considerations of this approach. The author advises treating access to Azure OpenAI Studio as an exception and mentions that Microsoft may make improvements over time. They also provide information on using debug mode or a proxy tool to examine API calls and highlight the need for specific data plane-level permissions. The statement also mentions the author's background and their blog, \"Journey Of The Geek,\" where they provide technical information on various technologies. #AzureOpenAI #APIKeyUsage #AzureActiveDirectory #softcorpremium

Review: CodeWhisperer, Bard, and Copilot X | InfoWorld

The article discusses various AI code generation tools, including GitHub's Copilot and Amazon's CodeWhisperer, which offer features such as code suggestions, pull request support, and security scanning. It also introduces Google's Bard and its capabilities in generating, debugging, and explaining code in over 20 programming languages. The author recommends using discretion and carefully testing and reviewing the generated code before using it. The article also provides tips for explaining code effectively and includes a sample dialogue between two individuals. #AIcodegeneration #GitHubCopilot #GoogleBard #softcorpremium

Load Geospatial data in BigQuery using DataFlow and GDAL | by Sandeep Manocha | Google Cloud - Community | Jul, 2023 | Medium

This article discusses the challenges faced in extracting and loading geospatial raster images into BigQuery using Google Cloud Dataflow, GDAL, and JQ. The author explains that the polygonization process using DataFlow GeoBeam was slow and not scalable, so they transitioned to the GDAL library. They also describe the issues with memory usage and the need to re-tile large raster files into smaller tiles. The article also discusses the challenges with BigQuery inserts for GeoJSON data and suggests alternative approaches. Overall, the article highlights the steps and solutions involved in processing geospatial raster images for analysis in BigQuery. #geospatialanalysis #BigQueryChallenges #rasterimageprocessing #softcorpremium

14 LLMs that aren't ChatGPT | InfoWorld

The article discusses the growing interest in artificial intelligence (AI) in the C-suite, particularly in relation to ChatGPT. However, there are numerous competitors entering the market, making it difficult to measure the quality of generative AI responses. The cost of building large language models is a major concern, but organizations like Meta and Microsoft have developed more cost-effective models such as Llama and Orca. Additionally, there are specialized chatbots and models like Jasper, Claude, and FrugalGPT optimized for specific tasks and cost savings. #AIinC-suite #GenerativeAI #CostEffectiveModels #softcorpremium

6 key features of SingleStore Kai for MongoDB | InfoWorld

SingleStore Kai is a new API that allows for real-time analytics on JSON data for MongoDB applications. It offers 100x faster analytical performance than most queries and supports vectors and fast vector similarity search. It requires no code changes, data transformations, or changes to existing queries, and supports the same MongoDB tools and drivers. SingleStore Kai offers the best of both worlds with the schema flexibility of a JSON document store and the speed and efficiency of a relational SQL database. Easy data replication is also included. #SingleStoreKai #RealTimeAnalytics #JSONData #MongoDB #Vectors #VectorSimilaritySearch #SQLDatabase #JSONDocumentStore #DataReplication #softcorpremium

AI’s impact on cost savings, productivity, and jobs | InfoWorld

The article argues that instead of using AI to eliminate jobs and cut costs, companies should focus on using AI to make their employees more productive. AI can be used to augment various job functions and boost productivity. The article suggests that the primary goal should be to raise productivity rather than lower costs. AI tools like GitHub Copilot can greatly improve the software development experience but cannot replace developers entirely. The article emphasizes the importance of maximizing the impact each employee can make and using AI to enable employees to do better work. It concludes by stating that AI is essential for increasing workforce productivity and urges employees to incorporate AI into their work. #AIforProductivity #MaximizingImpact #WorkforceProductivity #softcorpremium

Introducing new and updated models to Azure OpenAI Service - Microsoft Community Hub

Azure OpenAI Service has announced updates including the availability of new versions of gpt-35-turbo and gpt-4 models, price reductions for popular models, and enhanced steerability for the gpt-35-turbo model. The new gpt-35-turbo-16k model offers 4x the context length of its predecessor. Azure OpenAI Service has also made updates to the model lifecycle, allowing users to choose whether to auto-update to the default version or pin to a specific version. Data privacy is emphasized, with assurance that user data will not be used to improve models or shared with third parties. Users can apply for access to Azure OpenAI Service and join the GPT-4 waitlist. #AzureOpenAI #GPT4 #softcorpremium

AI is my copilot | InfoWorld

The use of Large Language Models (LLMs) and generative AI is transforming the role of software engineers. While these tools can streamline tedious tasks and automate the assembly of derivative software, they also have the potential to augment human capabilities in solving complex problems and crafting secure, sustainable solutions. However, training LLMs for quality is crucial and human expertise is still necessary to ensure quality and avoid the deployment of damaging, dangerous code. #LLMs #AI #softwareengineering #softcorpremium