Zhanxiang Huang

What if blockchain technologies are applicable to industries like healthcare and logistics? Professor Jianliang Xu, Head and Chair Professor (HKBU Department of Computer Science), and his team’s pioneering “Gas-efficient Merkle Index Tree” is slashing operational costs by almost 80% while enhancing data precision, paving the way for blockchain’s broader adoption in healthcare and beyond in Hong Kong. 𝐑𝐞𝐚𝐝 𝐌𝐨𝐫𝐞 (Chinese only): https://lnkd.in/ehqcC3Ey  🗝 Cluster: Data Analytics and Artificial Intelligence #Cluster_Data_Analytics_and_Artificial_Intelligence #HKBU #ResearchOffice #HKBUResearch #HKBURO #blockchain #BlockchainTechnology #DataAnalytics #interdisciplinary HKBU Faculty of Science

18

Huge release by Daekeun Kim and Hyo (HK) Choi from our AI GBB team in Korea, sharing Korean language benchmarks for our latest Azure OpenAI model, gpt-4o-mini - like gpt-4o, the model excels on Korean language tasks, in this case mini completely eclipses gpt-35-turbo and nearly achieves gpt-4-turbo performance. The benchmarks and code are open source, extensible beyond our Azure models, so we encourage our customers and the larger community to try it out and even fork/share your own findings - and please let me in the comments what other Asia languages beyond Korean you may be looking to evaluate with your language models.

52

7 Comments

Revolutionizing Patient Assistance Programs (PAP) with Automation and AI ! Patient Assistance Programs (PAP) play a critical role in supporting uninsured or underinsured patients, ensuring they receive essential medications and services. At PatientsForce, our MediPASS platform harnesses the power of automation and AI to streamline PAP operations, driving better patient outcomes and enhancing operational efficiency. Our automated solutions tackle key challenges in PAP management: AI-Powered OCR: Handling large volumes of diverse documents is a major pain point. Our platform uses AI-enhanced Optical Character Recognition (OCR) to swiftly extract and validate critical data, from medical records to application forms, reducing manual efforts and accelerating processing times. AI Voice Solutions: Traditional phone interactions are upgraded with AI-driven voice support, improving patient communication through real-time responses, standardized queries, and virtual agents. This ensures patients receive timely assistance, enhancing overall program engagement and compliance. Automated Eligibility Assessments: Determining patient eligibility requires intensive data collection and validation. Our automated system assesses patient data against predefined criteria, flagging inconsistencies instantly to streamline onboarding and re-enrollment. By integrating these solutions, PatientsForce optimizes PAP management, reducing costs, improving scalability, and ensuring patients receive the support they need in a timely and compliant manner. Automation is not just about efficiency; it’s about sustainability and delivering long-term patient care success. #HealthcareInnovation #Automation #AI #PAP #PatientsForce #MediPASS #PatientCare #HealthcareEfficiency

4

Demand for AI servers is strong. Fii, a subsidiary of Hon Hai, made nearly 40 billion in the first half of the year Hon Hai's subsidiary, Fii, reported strong financial results for the first half of 2024, driven by increased demand for AI servers. The company's net profit reached RMB 8.74 billion (approximately NT$39.8 billion), marking a 22.04% annual growth. Fii's operating income was RMB 266.09 billion (approximately NT$1.21 trillion), a 28.69% increase from the previous year. The growth is attributed to Fii's vertical integration capabilities and heightened revenue from AI server products, which have doubled. Fii plans to enhance its AI ecosystem, improve production capacity, and increase R&D investments to meet customer demands, focusing on GPU modules, substrates, and AI server design. #CloudComputing #Taiwan https://lnkd.in/gYXrT6AR

1

Hello LinkedIn connections, In light of the COMPUTEX TAIPEI excitement, I've written a two-part analysis on NVIDIA. This first part covers the company's founding background, business model, and key clients. 🙌 If you want to learn more, click the link below to read the full article: https://lnkd.in/g2FwefYa Here are 3 quick key takeaways: 1. NVIDIA's GPU Technology: NVIDIA is a market leader with its cutting-edge #GPU technology, especially in the #AI and gaming sectors. The GeForce series GPUs enhance gaming graphics quality, while Tensor cores dramatically accelerate the training and learning of AI models. The #CUDA platform also expands the range of GPU applications. 2. #AI Boom: With the surge in #generativeAI, NVIDIA's GPUs have become crucial for training AI models. Major tech companies like 微軟公司, 谷歌, and Amazon are all competing to use NVIDIA's GPUs to boost their AI capabilities. Additionally, thousands of AI startups rely on NVIDIA chips for fast and efficient training, which has significantly increased NVIDIA's market value. 3. NVIDIA's Business Model: NVIDIA has transitioned from just a GPU supplier to a comprehensive system provider. They offer complete AI solutions, including #Spectrum-X data center technology for training, the #Hopper training platform, and the #Blackwell operational platform, selling these as Total Solutions to businesses. NVIDIA has built a powerful platform ecosystem that attracts numerous developers and corporate clients. Their strategic use of leverage and scale is key to maintaining their competitive advantage. For more interesting stories about tech industry, subscribe to Roxanne’s Tech Talk👩🏻‍💻 I share in simple, understandable language, making technology accessible to everyone! Subscribe 👉 https://lnkd.in/gpzggdvb

42

3 Comments

DevFest Kaohsiung 2024 講者公開 🔗 活動報名 https://lihi.cc/WBTgI 💬 講者介紹 Jerry 老師『APMIC - Founder & CEO』 https://lnkd.in/geG5CEMv Jerry 老師是現任亞太智能機器 (APMIC) 的創辦人兼執行長，也是 Google 機器學習開發專家 (GDE)，專注在自然語言理解 (NLU)。在亞太智能機器，Jerry 帶領團隊開發了基於 LLM 解決方案 APMIC CaiGunn，提供企業級的 LLM 工具，該產品於 2023 年入選 NVIDIA 的 GenAI 軟體供應鏈，同步也於 2024 年入選亞太區 Google Cloud 的機器學習個案廠商，他致力於將 AI 工具化，讓人人都可以使用 AI 來創造價值。 💬 議程主題 ShieldGemma in RGAI 💬 議程內容 Google present ShieldGemma, a comprehensive suite of LLM-based safety content moderation models built upon Gemma2. These models provide robust, state-of-the-art predictions of safety risks across key harm types in both user input and LLM-generated output. 💬 議程標籤 #負責任的人工智慧 #ResponsibleAI #RGAI #Gemma #OpenModel 🕟 議程時間 PM 16:30 ︴A1 大演講廳 #DevFest #KOIN智高點

3

2 Comments

“At Akeso, we have diverse clinical teams that foster a culture of mutual learning. This collaborative atmosphere enables us to solve problems quickly and efficiently. When challenges arise, we can promptly adjust our strategies to ensure smooth research and development progress. Our management team is essential in identifying global trends, combining strategic vision with hands-on support for frontline staff to address issues swiftly. In today’s market, this approach is crucial for any innovative pharmaceutical company looking to enhance efficiency and maintain competitiveness.” Insights from the Forbes China interview with Michelle Xia, Founder, Chairwoman, President and CEO of Akeso. https://lnkd.in/g-UCqKM2

9

At the end of last year, AWS announced Graviton4 and Amazon EC2 R8g. We expected these two innovations will launch in 2024 H1, maybe March or April. Until previous week, we knew they finally made R8g general available now. And we start to trace back all the features they revealed before. Re:invent 2023 is first step. We look up relative breakout sessions video and blogs, press release. Second clue is re:Inforce 2024. We trace the security enhanced features of Gravition4, and like to know more about full encryption among high speed hardware interfaces and BTI. The AWS CISO mentioned about the newest custom-built CPU will also adopt Point Authentication and No SMT. https://lnkd.in/gN5Uqf6A

6

We are thrilled to announce the launch of our new master’s programs in 2025/26 intake: MSc in Information and Cyber Security Management (MScICSM) (official launch in early 2025): This innovative program incorporates the psychology behind cybercrime, which is not covered in typical traditional courses, rather than just technical skills. Our Acting Dean of the HKUST Business School, Prof. Kai-Lung Hui, highlighted that most current cybercrimes are unrelated to hardware or technical knowledge; instead, criminals often exploit criminal psychology to carry out fraud on online platforms. In this program, students will develop essential management competencies, covering risk management, criminal psychology and cybersecurity policies. It is designed to be accessible to all, regardless of technical background. MSc in Family Office and Family Business (MScFOFB) (open for applications): In collaboration with local family business heirs, this program will delve into family office culture and operations through real-world case studies. It's an opportunity for those looking to transition into this fast-growing field.   Stay tuned to our announcement on the new programs! 🚀

22

1 Comment

Possible Recommendations at the First Glance: 1. Enhance cloud security protocols: Regularly audit and update cloud storage infrastructure to prevent vulnerabilities. 2. Third-party risk management: Implement stricter controls and audits for third-party services and software to minimize security risks. 3. Proactive patching: Ensure timely updates to all critical systems to reduce exposure to known vulnerabilities. This incident serves as a reminder for organizations to stay vigilant against emerging threats and ensure their security posture is continuously evolving.

4

[Updated weekly🐍-Python-] Save data construction part 1 https://lnkd.in/gFfrckXs 🐣SILVER-API- scheduled to be released in May is now available🐣 Topic: Exporting and importing text files✍️ This time, save data construction - Python - Part 1! 👍 After exporting a large amount of data, just leave a note using this ReaScript 📓 You can easily check where and how many of what you have saved at any time. Please check the blog for details🎧 #reaper #Reascript #RIGDOCKS #REAPDOCK #APIDOCK

1

This week we are presenting our paper "LiNR: Model Based Neural Retrieval on GPUs at LinkedIn" accepted at CIKM 2024 (https://lnkd.in/gUrWqRcD). Please stop by and say hi to Aman Gupta, who will be there in person :) We discuss our experiences and challenges in creating scalable, differentiable search indexes using TensorFlow and PyTorch at production scale. In LiNR, both items and model weights are integrated into the model binary. Viewing index construction as a form of model training, we describe scaling our system for large indexes, incorporating full scans and efficient filtering. We believe LiNR represents one of the industry's first Live-updated model-based retrieval indexes at production scale. Talented co-authors include Fedor Borisyuk, Qingquan Song, Mingzhou Zhou, Ganesh Parameswaran, Madhulekha Arun, Siva P., Tugrul Bingol, Zhoutao Pei, Stanley(Kuang) Lee, Lu Z., Hugh Shao, Syed Ali Naqvi, Sen Zhou, Aman Gupta

169

4 Comments