News

Last Week in AI: Nokia's EDA Platform, Salesforce's Tableau Einstein, Uniphore's X-Stream, More

• By Pure AI Editors
• 09/23/2024

This edition of our roundup of AI products and services announced last week that didn't make headlines, but should be on your radar includes Nokia's EDA Platform, Salesforces' Tableau Einstein, Uniphore's X-Stream, and more.

Nokia launched its Event-Driven Automation (EDA) platform last week. Built on Kubernetes, the EDA was designed to reduce network disruptions and service downtime by minimizing human error in data center operations. The platform introduces a modern, intent-based approach to network automation and simplifies life cycle management, potentially cutting operational effort by up to 40%, the company claimed. The platform, available through on-prem and cloud-based subscription models, features digital twin capabilities, generative AI assistance, and integration with various IT service management systems and cloud platforms. With this platform, Nokia is aiming to address the growing demand for scalable, multi-vendor solutions amid workforce shortages and the rise of AI-driven workloads. Nokia will be hosting a live event this week (Sept. 24) for an in-depth look at EDA's innovations in data center networking.

Salesforce announced Tableau Einstein, a reimagined, AI-powered, visual analytics platform that includes Agentforce.  Salesforce Tableau is a data visualization and business intelligence (BI) platform designed to allows users to explore, visualize, and analyze data from multiple sources. Agentforce is a tool to build and manage autonomous agents for tasks across various business departments. The company called the new Tableau Einstein "a reimagined analytics experience that accelerates the path from raw data to insights by infusing autonomous and assistive agents into every aspect of analytics to help everyone answer, engage, and act on insights directly from where they work in real time." Because Tableau Einstein is built on the Salesforce Platform, any team can use analytical insights to make strategic decisions more quickly and effectively, the company said. With Tableau Einstein, data architects and analysts can build semantic models with agents integrated into the authoring flow, automatically applying recommended relationships between objects for context and augmenting definitions. A new dedicated marketplace will help developers create, reuse, and share apps and analytical assets more easily across the company and an extended Tableau Public. 

DevOps toolmaker JFrog announced the launch of JFrog Runtime, a new end-to-end security solution, along with a product integration with NVIDIA to secure AI models. JFrog Runtime was designed to provide real-time vulnerability detection and safeguards applications from development through production, addressing security risks in cloud-native environments like Kubernetes. It enhances collaboration between development and security teams by offering a unified platform for managing software lifecycle risks. The NVIDIA integration enables enterprises to deploy secure, GPU-optimized AI models through Nvidia NIM microservices, streamlining AI model management and improving performance. And it allows developers to quickly deploy compliant AI models in production, the company said.

Endevor, a provider of software for asset management and operational excellence, launched ENGAGE Intelligence, a new platform that integrates data science technologies into its ENGAGE suite. The tool aims to boost automation and efficiency for industries such as power generation, utilities, and manufacturing. The platform features machine learning, process optimization, and data visualization to streamline asset management. ENGAGE Intelligence offers tools for optimizing maintenance schedules and capital project planning, enabling better resource allocation. Additional features include data-rich visualization dashboards with drill-down capabilities, automated text generation using local large language models (LLMs), and text field revision tools.

Uniphore, a leading AI-native enterprise company, introduced X-Stream, a new platform capability designed to transform structured and unstructured multimodal data into domain-specific knowledge for enterprises. Positioned as a core layer within Uniphore's X-Platform, X-Stream enables knowledge-as-a-service, allowing businesses to create AI applications that enhance their digital transformation efforts. The platform addresses key challenges in AI adoption, including data quality and cost concerns, by offering advanced retrieval-augmented generation (RAG) capabilities, enterprise-scale knowledge graphs, and AI-ready data transformation. X-Stream’s features include tools to ensure output accuracy, mitigate bias, and provide robust governance and compliance, positioning it as a crucial asset in the growing race for AI dominance.

Qwen has announced the release of Qwen2.5, its latest addition to the Qwen family of language models, marking what the company claims could be the largest open-source release in history. The new models include Qwen2.5, specialized coding model Qwen2.5-Coder, and math-specific model Qwen2.5-Math. Available in various sizes, the models span from 0.5 billion to 72 billion parameters, with most models licensed under Apache 2.0. Qwen2.5 introduces significant improvements in language processing, coding, and mathematical capabilities. The models are designed to handle multilingual tasks and support long text generation, making them resilient to complex system prompts. The release also features APIs for flagship models Qwen-Plus and Qwen-Turbo, and enhanced performance for the 72B version of Qwen2-VL. This new generation of models is expected to unlock innovative uses in coding and mathematics, with Qwen touting its potential for significant contributions to the open-source AI community.

SiFive, provider of RISC-V-based processors and silicon chips, announced the SiFive Intelligence XM Series designed for accelerating high performance AI workloads. This is the first IP from SiFive to include a highly scalable AI matrix engine, the company said, which accelerates time to market for semiconductor companies building system on chip solutions for edge IoT, consumer devices, next generation electric and/or autonomous vehicles, data centers, and beyond. SiFive’s new XM Series offers an extremely scalable and efficient AI compute engine. By integrating scalar, vector, and matrix engines, XM Series customers can take advantage of very efficient memory bandwidth. The XM Series also continues SiFive’s legacy of offering extremely high performance per watt for compute-intensive applications. The company also announced plans to open source a reference implementation of its SiFive Kernel Library (SKL).

Kurve, a venture-backed startup building an AI-powered developer tool that automates data discovery is coming out of stealth and launching its beta product. The company, a spin off from SRI International, "is set to redefine how enterprises extract value from their ever-expanding data estates," said Wes Madrigal, Co-Founder and CEO of Kurve, in a statement. Kurve’s product is designed to learn relationships between datasets at the column level and extend these insights to the semantic level, automating the creation and maintenance of metadata ontologies. This approach removes one of the largest obstacles in data science workflows: platform incompatibilities. And it significantly accelerates the processes of data discovery, understanding, and feature engineering, which enables businesses to unlock the full potential of their data more efficiently than ever before.