​AI Tech Circle​

Stay Ahead in AI with Weekly AI Roundup; read and listen on AITechCircle:

Welcome to the weekly AI Newsletter. This newsletter is a valuable resource for me and many others in the community. It provides practical and actionable ideas that can be immediately applied to our jobs and businesses.

This week, I am keeping the focus on the CloudWorld updates.

Before we begin, please share this week’s updates with a friend or colleague:

Today at a Glance:

• New announcements from CloudWorld
• AI Weekly news and updates covering newly released LLMs

Oracle CloudWorld 2024 Recap: What You Need to Know

​

The multi-cloud and AI announcements dominated Oracle’s flagship annual conference this week. Today’s edition covers all essential products and features announced during the CloudWorld 2024.

Let’s directly review the different areas:

AI:

• OCI Generative AI Agents with retrieval-augmented generation (RAG) capabilities to help customers more easily apply AI to real-world business operations and gain a competitive advantage from their data. The RAG Agent provides out-of-the-box RAG capabilities, enabling customers to start quickly without manual processes like agent planning, retrieval, reranking, generation, and integration. It includes a self-check feature to reduce AI hallucinations, allowing businesses to adopt RAG technologies to streamline processes without extensive research and development. OCI GenAI Agents also grant access to Oracle Database 23ai AI Vector Search for fast similarity queries on enterprise data stored in the database.
• Generative development (GenDev) for enterprises is an AI-centric application development infrastructure. GenDev offers innovative technologies enabling developers to rapidly generate sophisticated applications and integrate AI-powered natural language interfaces and human-centric data. It combines features from Oracle Database 23ai, including JSON Relational Duality Views, AI Vector Search, and APEX—to facilitate generative AI development.
• Select AI Synthetic Data Generation (SDG) to use LLMs to generate Synthetic, or artificially generated, data conforming to your schema for solution testing, proofs of concept, and other uses.
• AI Agents in SaaS – ERP Fusion applications:
  • Ledger Agent: Eliminates manual effort by detecting exceptions and anomalies in transaction data and efficiently monitors account balances. Provides detailed sub-ledger insights through natural language queries. It can identify revenue variances before quarter-end, investigate factors like order delays, and automate necessary journal entries.
  • Document IO Agent: Automates onboarding of complex third-party integrations, enhancing efficiency in document capture and generation across various transactions, channels, standards, formats, and languages. It ingests and standardizes diverse documents like images and PDFs, converting them into requisitions, invoices, or payment instructions for review and approval.
  • Advanced Prediction Agent: Supports multivariate AI prediction models by leveraging financial, operational, and external factors in predictive forecasting. Creates data-driven revenue forecasts by uncovering hidden patterns and trends, resulting in more timely and accurate cash forecasts.
• AI Agents in SaaS – HCM applications:
  • Shift Scheduling Assistant: Optimizes employee shift schedules by accommodating individual preferences and ensuring regulatory compliance. Assists schedulers in understanding overscheduling policies and regulatory implications.
  • Employee Hiring Advisor: Streamlines candidate sourcing with optimized campaigns, helping hiring managers and recruiters create requisitions and offers to reduce hiring time. Aids in completing requests for new positions, filling roles, and developing job offers that align with company policies.
  • Benefits Analyst: Simplifies access to employee benefits, enabling employees to understand and optimize their benefit packages based on individual needs. Provides clarity on benefits included in different packages and compares options to support informed decision-making.
• AI Agents in SaaS – SCM applications:
  • Customer Sales Representative Guide: Enhances customer experience by providing personalized insights and recommendations for handling order queries. Assists in understanding customer service policies regarding delays or defects and advises on how to notify customers about impacts to their orders.
  • Maintenance Troubleshooting Advisor: Accelerates maintenance and repairs by offering personalized insights and recommendations for asset upkeep. Provides guidance from equipment manuals on troubleshooting topics like error codes, possible causes, and repair steps.
• AI Agents in SaaS – CX applications:
  • Customer Account-Researcher Agent: Provides sales teams with automated insights for planning and research tasks, freeing up time to build relationships and drive account growth. It summarizes account health, identifies upselling and cross-selling opportunities, and highlights critical stakeholders and corporate initiatives.
  • Contracts Researcher Agent: This agent automates routine contract workflows and approvals, allowing sales teams to focus on selling instead of administrative tasks. It streamlines contract authoring and renewals while ensuring compliance with vendor best practices.
  • Incentive Compensation Plan Guide: Helps organizations motivate sales representatives by clearly communicating compensation plans aligned with strategic goals. Sellers can access clear explanations of their compensation plans to maximize earning potential throughout the sales process.

Multi-Cloud:

• ​Oracle & AWS have jointly announced the launch of Oracle Database@AWS. This new service allows customers to access Oracle Autonomous Database and Oracle Exadata Database Service directly within the AWS environment. This offering provides a unified experience between Oracle Cloud Infrastructure (OCI) and AWS, simplifying database administration, billing processes, and customer support. Customers can seamlessly connect their enterprise data stored in Oracle Databases to applications running on Amazon Elastic Compute Cloud (Amazon EC2), AWS Analytics services, and advanced AI and machine learning services like Amazon Bedrock.
• ​Oracle & Google Cloud have announced the general availability of Oracle Database@Google Cloud in four regions across the United States and Europe. Customers can now run Oracle Exadata Database Service, Oracle Autonomous Database, and Zero Data Loss Autonomous Recovery Service on Oracle Cloud Infrastructure within Google Cloud data centers.

OCI Infrastructure & PaaS

• Oracle Cloud Infrastructure (OCI) is now accepting orders for the most prominent AI supercomputer in the cloud—the world’s first zettascale computing cluster. Featuring up to 131,072 NVIDIA Blackwell GPUs, it delivers an unprecedented 2.4 zettaFLOPS peak performance.
• Oracle is introducing Dedicated Region25, a new OCI Dedicated Region configuration that is smaller, scalable, and deployable within weeks, starting with just three racks. With a 75% smaller launch footprint and simplified data center requirements, Dedicated Region25 supports over 150 of OCI’s AI and cloud services. This allows more customers to gain the agility, economics, and scale of the public cloud within their own data centers. The new configuration will be available next calendar year.
• OCI Zero Trust Packet Routing (ZPR) is integrated into OCI’s network fabric. This feature enhances data security by separating network security from the underlying architecture, helping to prevent unauthorized data access. OCI ZPR allows organizations to set security attributes on resources and write natural language policies to limit network traffic based on accessed resources and data services. This helps safeguard against common causes of compromise like network misconfigurations.
• Oracle Code Assist is an AI code assistant that helps increase developer productivity. Deployed as a plugin for JetBrains IntelliJ IDEA or Microsoft Visual Studio Code, Oracle Code Assist will provide developers with intelligent suggestions to help them build and optimize applications written in modern programming languages, including Java, Python, JavaScript, SuiteScript, Rust, Ruby, Go, PL/SQL, C#, and C.
• Oracle Intelligent Data Lake as a foundational component of the Oracle Data Intelligence Platform. This enables organizations to leverage powerful generative AI features to streamline workflows, enhance productivity, simplify code generation, provide conversational analytics, and create dashboards with minimal technical expertise.
  ​Oracle Analytics Cloud AI Assistant translates natural language into actions, bridging the gap between an analyst’s vision and its realization. It features a built-in large language model (LLM) optimized for analytics conversations and tasks, understanding the context of user queries, and recognizing Oracle Analytics workbooks and datasets.
• HeatWave GenAI offers integrated, automated, and secure generative AI capabilities, enabling developers to build new generative AI applications without needing AI expertise, data movement, or additional costs. LLM Inference Batch Processing enhances application throughput by processing multiple requests simultaneously across the HeatWave cluster.
• Robotic Process Automation (RPA) is now available as a native capability of Oracle Integration Cloud; further details​

Industry Apps

• Oracle has released a new RFID-based replenishment solution within Oracle Fusion Cloud Supply Chain & Manufacturing (SCM) to help healthcare customers optimize inventory management. The RFID for Replenishment solution, part of Oracle Fusion Cloud Inventory Management, utilizes RFID technologies from Avery Dennison, Terso Solutions, and Zebra Technologies to automatically capture usage, update stock balances, track locations, and trigger restocking of supplies and materials.
• New AI capabilities in Oracle’s revenue transformation solution include
  • Generative AI for Contract Summarization: Speeds contract processing and deal closures by quickly answering contract-related questions. It helps sellers and buying groups track key milestones and obligations, enhancing internal oversight.
  • Generative AI for Sellers: Enables sellers to quickly seize sales opportunities and engage buyers more effectively to drive account expansion. It assists in drafting emails, activity summaries, sales executive briefings, and executive summaries for quotes and proposals.
• Oracle Energy and Water Data Intelligence, a data unification, analytics, and AI solution designed for utilities. It integrates data from Oracle and third-party sources, providing utilities with a comprehensive view of their operations for advanced analysis and reporting. Business leaders can make critical decisions faster with access to actionable, pre-built insights. Data scientists can also extend the solution using self-service machine learning to create custom applications tailored to each utility’s needs, such as forecasting load growth in certain areas or recommending consumer actions based on utility programs.

Weekly AI News & Updates…

Last week’s AI breakthroughs marked another leap forward in the tech revolution.

1. OpenAI o1-mini and OpenAI o1: a series of new models from OpenAI. These models spend more time thinking (I would say more time ‘processing’) before responding, much like human thought processes. This helps them refine their strategies, try different approaches, and recognize mistakes. In testing, the latest model performs comparably to PhD students on challenging physics, chemistry, and biology benchmarks. It also excels in math and coding, scoring 83% on a qualifying exam for the International Mathematics Olympiad (IMO), whereas GPT-4 only solved 13% of the problems. In coding competitions, it reached the 89th percentile in Codeforces contests. In my practice for coding, it has improved a lot. link​
2. DataGemma from Google to address the hallucination problem in the LLMs. It will expand the capabilities of Gemma models by relying on the knowledge of Data Commons to enhance LLM factuality and reasoning using two distinct approaches. RIG (Retrieval-Interleaved Generation) enhances the capabilities of our language model, and RAG (Retrieval-Augmented Generation) enables language models to incorporate relevant information beyond their training data, absorb more context, and help more comprehensive and informative outputs. link​
3. NotebookLM! Create a notebook and upload one or more sources (e.g., PDFs of research papers, your favorite PhD thesis, a newspaper article, etc.); it will create a podcast of two voices discussing the content. I will try this in my podcast. link​
4. WorldLabs: a spatial intelligence startup from Dr. Fei-Fei Li, building Large World Models (LWMs) to perceive, generate, and interact with the 3D world. link​
5. Groq has introduced a partnership with Aramco Digital to establish the world’s largest inference data center using Groq LPU AI inference technology. link​

Download Generative AI Use Cases 100+ …

Use cases with Generative AI for you to map in your organization for implementation.

Earlier week’s Post:

Thank you for taking the time to read. Please don’t hesitate to hit reply and share what you found most helpful this week.

Your feedback is highly appreciated!

Until next week,

Kashif Manzoor