Brentley Beerline

Microsoft is powering next-level media solutions. MediaKind trusts Microsoft Azure as the backbone for instant, global distribution of content. By leveraging Azure’s scalable infrastructure, they can handle real-time coverage demands without missing a beat. Check out the story here: https://lnkd.in/gqxZnJzw Erik Ramberg, Allen Broome, Simon Crownshaw, Silvia Candiani, Jessica Alderson, Rich Anderson, MBA, Sebastian Rasino, Monica Reynolds, Kyle Nielsen, Eliot Sakhartov #MediaTech #MediaInnovation #BroadcastSolutions

93

1 Comment

AI is no longer experimental - it’s changing how we work, how decisions are made, and how businesses compete. But here’s what I hear from customers: it’s not the models that slow things down. It’s the data. Getting it ready, keeping it flexible, and scaling it without breaking things is a real challenge. That’s why we’re expanding our Enterprise Data Cloud (EDC) capabilities to address these challenges head-on. Our new Cloud Azure Native gives customers the freedom to run AI workloads where it makes the most sense. Portworx AI Copilot helps teams focus on meaningful outcomes instead of manual work. Together with NVIDIA, we’re making it faster to move AI from pilot to production. And because resilience isn’t optional, we’ve built cyber protection right into the storage layer - with built-in detection, automated recovery, and deeper partnerships with CrowdStrike, Superna, and Veeam Software to help keep mission-critical data safe. With EDC, our goal is simple: help enterprises scale AI with confidence - smarter, stronger, more resilient. Read more here: https://lnkd.in/g8uqNGtq, https://lnkd.in/gnhVSVbe #PureStorage #DataStorage #Partnerships #Veeam #Superna #Crowdstrike #MicrosoftAzure #NVIDIA #Innovation #Data #VMware

75

Today, we announced our intent to acquire the Spot by NetApp FinOps business from NetApp.   AI is a board-level discussion for CIOs and CTOs. Optimizing cloud footprints, that continue to expand, better enables businesses to shift more of their technology budgets to other strategic initiatives.   The combination of Flexera and Spot deepens our ability to tackle the proliferation of hybrid cloud cost and usage challenges across public cloud, data centers, SaaS, containers and even AI. There is no doubt this is the direction the market is headed in, and as the leader in technology spend and risk management, we’re leading the way here, too. Flexera and Spot will create a formidable offering for enterprises and MSPs, and this deal validates our commitment to and leadership in the FinOps market.   Learn more in our announcement: https://lnkd.in/gKpkecr7   #flexera #finops #spot

507

20 Comments

🧐 For those working on MCP servers: have you found that too many fine-grained tools don't perform as well as fewer, broad-functionality coarse grained tools? Do the LLMs that decide when/what tools to call, is it too much to rely on them to think like an imperative programming language? Rather, it would prefer to think in terms of broader functionalities? Let me give you a concrete example: Say the task is “summarize the status of a customer.” With fine-grained tools, the LLM might need to plan and execute something like: 1️⃣ Call getCustomerProfile(customerId) 2️⃣ Call getOrders(customerId) 3️⃣ Call getPayments(customerId) 4️⃣ Call getSupportTickets(customerId) 5️⃣ Correlate all the results 6️⃣ Build a summary That’s a lot of steps, and each one has failure or reasoning risk. Now imagine a coarse-grained tool: 👉 getCustomerContext(customerId) → returns profile, orders, payments, and tickets in one response. The LLM’s reasoning path simplifies to: 1️⃣ Call getCustomerContext(customerId) 2️⃣ Summarize Much less planning, fewer chances for error, and a reasoning pattern that aligns more with how LLMs prefer to work (broad, declarative intent) rather than imperative step-by-step coding. What are others seeing?

43

33 Comments