Nexus Connect provides a single, university-branded connector inside AI chat agents, giving students and researchers seamless access to scholarly resources and services.

LONDON, April 28, 2026 /PRNewswire/ — Clarivate Plc (NYSE: CLVT), a leading global provider of transformative intelligence, today announced Nexus Connect, an AI gateway to universities’ academic resources and services. Nexus Connect is designed to meet users wherever they work — as a connector inside leading general-purpose AI chat agents, such as ChatGPT or Claude.

AI chat agents increasingly serve as the main workspace for students and researchers. In recent months, Model Context Protocol (MCP) — an open standard that lets AI chat agents connect directly to external data sources, tools and services — has emerged as a powerful way to bring academic resources directly into these environments and ground AI responses in trusted content. But as vendors across the academic ecosystem release product-specific MCP services, libraries are left out of the picture — with no visibility, control, or certainty that their licensed resources and services are surfaced.

Nexus Connect addresses this directly, integrating with other platforms that support MCP integration. Powered by the Clarivate Academic AI Platform, it gives users easy access to their university licensed content and services, directly from AI chat agents. It brings together Clarivate content and services, library resources, and services from other academic vendors into a single institutional presence. Nexus Connect joins Clarivate Nexus Extend, a browser-based academic assistant, as part of Clarivate’s growing suite of solutions designed to meet AI users where they are.

Oren Beit-Arie, Senior Vice President, Strategy & Innovation, Academia & Government at Clarivate said: “Libraries are where trusted knowledge lives, and our goal is to make sure that remains true in AI environments.

“Nexus Connect is the latest example of our strategic approach to academic AI – embedding transformative intelligence into our products, containing our carefully curated data and content, while connecting AI users to trusted resources. We are providing institutions with a unified layer that keeps their resources and identity in front of students and researchers. Alongside Nexus Extend, it keeps libraries at the center of research and learning.”

Shirley Wong, University Librarian, The Hong Kong Polytechnic University said: “User expectations are changing fast as they want the speed and ease of AI as part of how they study and research. Our priority is to bring AI to users in a way they can trust, grounded in the library’s resources and expertise. MCP lets us do that, by building trusted AI services within our institutional environment.”

Five universities in North America and APAC will be among the first institutions to deploy Nexus Connect. A broad early access program is planned for July 2026.

Core capabilities include:

Entitlement-based access: Users can discover the resources their library has licensedInstitutional identity: The library stays visible and central, under the university’s brandConfigurable scope: Libraries control which services are exposed and how content sources are prioritizedExtensible architecture: Institutions can start with one service and expand over timeEnterprise-grade security and privacy: Built to meet security and privacy standards

Nexus Connect launches with two use cases available to early adopter partners:

Academic discovery: powered by Primo and the Clarivate Central Discovery Index (CDI), students and researchers can search and retrieve library materials – including local resources – directly from their AI chat environment, with links into the library’s holdings for full access.End-user library services: powered by Alma and Primo, users can manage book loans, holds, renewals and library accounts directly from their AI chat agent.

In practice, this means a student working in ChatGPT or Claude can search their university’s entire library collection, find an e-book on their topic and link to the full text, all without leaving the chat. If a print copy is needed instead, they can check availability and place a hold in the same conversation.

Subsequent releases will expand Nexus Connect with additional resources and services, including library expert resources, such as research guides and teaching materials; additional Clarivate products; and services from other vendors across the academic ecosystem.

Clarivate works with leading general-purpose AI chat agent providers to deliver purpose-built institutional connectors that fit the needs of academic institutions — addressing priorities such as ease of deployment, usage attribution, and copyright protection.

Nexus Connect prioritizes security and privacy, to meet the requirements of academic institutions. It operates in accordance with Clarivate’s privacy policy and applicable data privacy regulations. 

Libraries and universities who wish to shape the Nexus Connect roadmap can apply for the early access program.

About Clarivate  
Clarivate is a leading global provider of transformative intelligence. We offer enriched data, insights & analytics, workflow solutions and expert services in the areas of Academia & Government, Intellectual Property and Life Sciences & Healthcare. For more information, please visit www.clarivate.com.  

Media Contact:  

Amy Bourke-Waite, Senior Director, External Communications 
newsroom@clarivate.com  

Logo – https://mma.prnewswire.com/media/1159266/5939570/Clarivate_Logo_v1.jpg

View original content:https://www.prnewswire.co.uk/news-releases/clarivate-introduces-nexus-connect-the-first-institutional-ai-gateway-to-trusted-research-and-learning-302755057.html

KYOTO, Japan, April 28, 2026 /PRNewswire/ — ROHM Co., Ltd. has developed a wireless power supply IC chipset consisting of the receiver (ML7670) and transmitter (ML7671) compatible with Near Field Communication (NFC) technology for compact wearables such as smart rings and smart bands.

ML767x chip set: https://www.rohm.com/products/power-management/wireless-power?page=1&SearchWord=ml767#parametricSearch 

Figures: Product features
https://cdn.kyodonewsprwire.jp/prwfile/release/M106254/202604217802/_prw_PI1fl_5G9aZOC5.jpg 

The smart ring market has seen rapid growth in recent years. However, for extremely small ring-shaped devices worn on the finger, wired charging is impractical, while the conventional Qi wireless charging standard is difficult to implement due to constraints such as coil size.

In response, NFC-based charging, which operates at the high-frequency 13.56MHz band that enables antenna miniaturization, is attracting increased attention, with adoption accelerating in next-generation wearables.

This new chipset builds on the proven receiver (ML7660) and transmitter (ML7661). The maximum power transfer is specified at 250mW, while peripheral components such as the switching MOSFETs required to supply power to the charging IC are built in. The result is a solution optimized for both mounting area and power transfer efficiency in the power class demanded by compact wearable devices.

Figures: Comparison of power receiver IC features and application examples
https://cdn.kyodonewsprwire.jp/prwfile/release/M106254/202604217802/_prw_PI2fl_t3V77HE4.jpg 

The ML7670 power receiver IC achieves a maximum power transfer efficiency of 45% in the 250mW low output range — all in an industry-leading form factor of just 2.28 x 2.56 x 0.48mm. What’s more, all firmware required for wireless power delivery is embedded directly within the IC, eliminating the need for a host MCU.

Compliance with NFC Forum (WLC 2.0) enables power transfer while maintaining compatibility with existing devices, positioning the chipset as a core element in the expanding NFC wireless power ecosystem.

Figures: Lineup table
https://cdn.kyodonewsprwire.jp/prwfile/release/M106254/202604217802/_prw_PI3fl_3aWk93R6.jpg 

The new chipset is already in mass production. Furthermore, it has been adopted in SOXAI RING 2, the latest model launched on December 10, 2025, by SOXAI Inc., the Japanese developer and distributor of the original sleep-monitoring ring SOXAI RING.

*The name “SOXAI” is pronounced “SOK-sai.”

SOXAI RING 2 Adoption Example: https://www.rohm.com/collaboration/soxai_ring-2 

Application Examples
– Smart rings
– Smart bands
– Smart pens
– Wireless earphones
– Other compact devices (i.e., wearables)

News release: https://www.rohm.com/news-detail?news-title=2026-04-28_news_wireless-charge&defaultGroupId=false 

About ROHM: https://kyodonewsprwire.jp/attach/202604217802-O1-7Qv27crN.pdf 

Logo: https://cdn.kyodonewsprwire.jp/prwfile/release/M106254/202604217802/_prw_PI4fl_2aE455Sk.jpg 

Official website: https://www.rohm.com/ 

View original content:https://www.prnewswire.com/news-releases/rohm-launches-ultra-compact-wireless-power-chipset-for-wearables-302754997.html

SOURCE ROHM Co., Ltd.

SYDNEY, April 28, 2026 /PRNewswire/ — Saigon Technology, a global software development and AI engineering company working with enterprises across Australia and APAC, recently reveals the biggest barrier to scaling AI is no longer model performance. It is the ability to connect predictive and generative systems into a single, production-ready architecture.

Australian enterprises are accelerating their investment in artificial intelligence. Organizations are deploying both predictive models to forecast outcomes and generative AI to create content, automate workflows, and improve customer engagement.

But a critical challenge is emerging.

Hybrid AI: The Next Strategic Imperative

Hybrid AI, which combines predictive and generative capabilities, is increasingly seen as the next step in enterprise AI adoption. Predictive AI identifies what is likely to happen. While generative AI determines what should happen next. When integrated effectively, the two create a closed loop between insight and action.

For example, a predictive model may identify a customer at risk of churning. A generative system can then produce a personalized retention offer tailored to that customer’s behavior and preferences.

This is where business value is created. It is also where most implementations fail.

The Integration Gap

A common pattern appears across enterprise AI projects.

Organizations develop predictive and generative systems separately, often with different teams, data pipelines, and success metrics. Then they try to integrate them at the end.

“It sounds practical, but it is the root cause of most hybrid AI failures.” said Thanh Pham, CEO of Saigon Technology

The generative AI team optimizes for output quality. The predictive team focuses on model accuracy. Neither is responsible for how the systems interact. Connecting them at the end often consumes months of rework.

Saigon Technology estimates that 60-70% of hybrid AI project costs are spent on integration, not model development, which is avoidable. So, integration architecture is now the first thing they design, before either model gets built.

Why AI Projects Stall Before Production

Integration is only one part of the challenge.

Beyond integration, three factors that prevent AI initiatives from reaching production:

Data readiness is underestimated: Most effort goes into cleaning, structuring, and maintaining data pipelines.The prototype-to-production gap is ignored: Models that perform well in testing often fail under real-world conditions without rigorous stress testing.User adoption is not addressed early enough. Even good AI can fail if business users do not trust or understand the outputs. Adoption, therefore, becomes as critical as development. 

A Framework for Scalable AI

To address these challenges, Saigon Technology has developed a 4-Layer Hybrid AI Framework, designed for production-scale deployment.

The framework includes:

A unified data foundationA model coordination layerBuilt-in governance for accuracy and complianceContinuous feedback loops to improve performance over timeThis approach has been applied for hundreds of Saigon Technology’s worldwide clients across industries.

Looking Ahead

As AI adoption accelerates across Australia, the gap between experimentation and production is becoming more visible. The succeed companies will be those design systems that work together from the outset.

Saigon Technology continues to partner with enterprises across Australia, the US, and APAC to build integrated AI systems that move beyond isolated use cases and deliver measurable business impact.

For organizations planning their next AI initiative, the message is clear: Integration is not the final step. It is the starting point.

Learn more at saigontechnology.com

View original content to download multimedia:https://www.prnewswire.com/apac/news-releases/saigon-technology-highlights-the-1-mistake-in-combining-predictive-and-generative-ai-302755348.html

SOURCE Saigon Technology