Clemens Vasters

Struggling to utilize 200K token context window of Anthropic Claude v2.1 & above models? Nampreet Sarao and I have just published an article on how to solve for this. Pls give it a read. https://lnkd.in/eHWcCHeJ #GenAI #AWSBedrock #AnthropicClaude #ContextWindow

52

2 Kommentare

Join us at Google Cloud Summit Berlin on May 29th! Google Cloud Security team will be on-site to discuss how Google Cloud's advanced security features can empower you to proactively identify and neutralize threats. Don't miss our security live demos about: - Threat Intelligence and Expertise: to understand threat actor activities and take actions on the threats that are most relevant to your industry - Security Operations : a modern, AI and Intelligence-driven SecOps platform that empowers your security teams to better defend against today’s and tomorrow’s threats - Cloud Security : to help you protect your cloud workloads and meet compliance & sovereignty requirements Let's talk about how we can make Google part of your security team! Have a look at the complete security agenda “The new way to protect” here ⬇️

2

At #T3DD24, I talked about the architecture of the mittwald cloud platform, the operational challenges that we encountered, and our future plans for ensuring long-term reliability and scalability. I've finally gotten around to posting a (two-part) write-up of said talk on the mittwald blog, brought up-to-date with everything that's changed in the past few months and a few more words on our plans for the very immediate future (in german; translation provided by the AI assistant of your choice). https://lnkd.in/gTpQBEwb

22

3 Kommentare

Orchestration services in SAP's generative AI Hub, SAP AI Core, is a true game-changer. Never before was access to LLMs this business-ready, with all the three Rs -> Reliable, Relevant, Responsible. #SAPBusinesAI

12

AI-generated code can behave in unexpected ways if developers don't thoroughly review it! Dynatrace CTO Bernd Greifender explains six key reasons why AI-generated code could make it more difficult to maintain the reliability and security of software over the long term

5

🚨Platform Health and Problem Reporting Template🚨 Kudos to Wolfgang Beer who in a recent post showed some new visualization options in our Dynatrace dashboards - such as honeycomb. But - his dashboard is a great template for #infrastructure and #platform owners to understand, report and tackle problem hotspots that jeopardize the resiliency of your infrastructure and applications. 📺My short video here gives you an overview of that dashboard! 📩You can also access and download the dashboard directly from the Playground tenant: https://lnkd.in/dvMRQ9UM ⁉Got questions or feedback: Just leave a comment here or send a DM! #observability #infrastructure #application #dynatrace #playground #template

83

3 Kommentare

Bring AI to life with DataStax, NVIDIA, and Wikimedia Deutschland e. V.! Catch the session replay from #AWSreInvent and learn how you can reduce AI development time by up to 60%. ⏱ Hear how two developers were able to ingest, chunk, and vectorize one of the world’s largest and most complex datasets in just three days (Much faster than other solutions) 🚀

9

Big news for Azure OpenAI Service, we are rolling out: ✅ Self-service provisioning and model independent quota requests ✅ Visibility to service capacity and availability ✅ Provisioned hourly pricing and provisioned reservations Read more on the tech community blog!

48

"Unlocking the Power of SVMs: My Personal Take on the Game-Changers in Machine Learning! 🚀 #MachineLearning #SVM" Ever thought machine learning was just for the tech elite? Think again! Support Vector Machines (SVMs) might sound like rocket science, but they're actually your secret weapon in decoding complex data patterns. Here’s why I’m all in: Beyond Linear: While logistic regression plays it straight, SVMs thrive on the curves. They're the masters of nonlinear classification, finding patterns where others see chaos. 🌀 The Art of Separation: Imagine separating apples from oranges, not with a line, but with a highway. SVMs find the widest path to distinguish data points, reducing errors like a boss. 🍎 vs. 🍊 Versatility is Key: From recognizing faces to predicting stock movements, SVMs adapt across challenges. Their flexibility in solving diverse problems is nothing short of impressive. 🌟 Mathematical Magic: At their core, SVMs are like puzzle solvers, using quadratic programming and the genius kernel trick to operate in high-dimensional spaces without breaking a sweat. It’s math, but not as we know it. ✨ Adapting to Imperfection: The real world is messy, and SVMs get that. With soft margins and slack variables, they find balance in chaos, proving that perfection isn't everything. ⚖️ SVMs aren't just another algorithm; they're a testament to how thinking differently can unlock new potentials in data analysis. The blend of mathematical elegance and practical adaptability makes them stand out in the crowded world of machine learning. Curious to learn how SVMs can transform your data insights? Dive deeper into the science behind these fascinating tools and discover how to leverage their power in your next project. #DataScience #Innovation Who else is fascinated by the unseen forces shaping our data-driven world? Share your thoughts below! #TechTalk

1

AI-generated code can behave in unexpected ways if developers don't thoroughly review it! Dynatrace CTO Bernd Greifender explains six key reasons why AI-generated code could make it more difficult to maintain the reliability and security of software over the long term

2

Final preparations for our Data Intelligence Day, taking place this Thursday at The Square in Zurich, are in full swing! We are especially excited about the keynote "AI & Data Trends" by Dr. Carsten Bange (BARC) and the insightful Best Practices on Microsoft, Qlik, and iVIEW. Don't miss your last chance to register for free: https://lnkd.in/dMYf78GC #iDID #DataIntelligence #Qlik #Microsoft #iVIEW #Barc #DataTrends #DataAnalytics #AI

7

Nemotron 70B scored an impressive 85 on the Arena Hard benchmark, leaving GPT-4 and Claude Sonnet behind. Whether you're accessing it through HuggingChat or testing it locally, this model is making waves in the AI community. Let's see how it handles Python coding, ethical decision-making, and more! Key Features: 1- Superior Performance: Beats Claude 3.5 and GPT-4 on multiple benchmarks. 2- Accessible: Use Nemotron on HuggingChat or install locally with ease. 3- Diverse Capabilities: Excels in coding, reasoning, ethics, and more.

5

Low-Code No-Code is a mega trend in IT industry. Not only in Applications Development, but also in others like #MachineLearning. Ludwig, a popular open-source low-code framework for building custom AI models like LLMs and other deep neural networks, it simplifies the machine learning process with a low-code approach, enabling non-ML roles(application developers, business users etc) and empowering the ML professionals(ML Engineers, Data Scientists etc) to quickly experiment, build, train, and deploy models When paired with SAP AI Core, these machine learning operations can be streamlined and managed at scale, fostering faster AI-driven innovation across the enterprise with enterprise-grade of security, scalability and compliance. The benefits of combing two together. ✅ Accelerated innovation of enterprise AI ✅ Improved Accessibility and Productivity ✅ Enterprise-grade scalability, security, and compliance ✅ Seamless Integration of Custom Intelligence into Business Process Here is my exploration with end-to-end Low-Code Machine Learning with Ludwig and SAP AI Core. I am able to replicate some ML models like Book Genre Classification, Defect Detection with image segmentation, and predictive maintenance based on machinery sound anomaly detection in hours. The only code is just a declarative configuration file in yaml. Check it out. Gianluigi Bagnoli Jacob T. Mostafa Sharaf Alice Magnani Cesare Calabria Trinidad Martinez Thiago Mendes Dr. Merza Klaghstan Alessandro Biagi Pavel Penaz Peter von Linstow

30

2 Kommentare

Learn how Run:ai on #OCI can help organizations scale their #AI capabilities without corresponding increases in operational complexities or costs. https://lnkd.in/dmF4FKkq #AI #Kubernetes

1