Operations | Monitoring | ITSM | DevOps | Cloud

LLM observability is key to ensuring model performance. Learn its importance, best practices, and actionable steps for optimal results and reliability.

Happy AWS re:Invent week, everyone! While we have some larger announcements coming out this week (keep an eye out), we’d be remiss if we didn’t take the time to also look back at what our teams have been working on over the past few weeks. While we didn’t have a stated theme for November, you will see that two of our updates revolve around building templates or examples to help make achieving your goals easier with CloudZero.

I recently had a conversation with one of our customers at AWS re:Invent that made me rethink the true value of automation. This customer told me, “You’ve automated so much for us that we almost forget about Qovery. Nobody even asks how to do things anymore.” That statement stuck with me.

High-performance computing (HPC) configuration has become paramount for AI workloads. Companies investing in AI workloads often find themselves with predictable problems: Lots of nodes that are hard to integrate and harder to manage together, under-utilized resources, deployment struggles, inconsistent configurations, and drift.

We’ve got another big update! This time it’s all about PostgreSQL; the entire product line has been polished and augmented to make your experience with PostgreSQL fuller in every way—this includes new data sources, improved suggestions, a slew of new smart data generators, and quite a few other tricks to make your daily work easy as ever. Without further ado, let’s check everything feature by feature!

Here comes the long-awaited release of three new Python Connectors for Dynamics 365 Business Central, Excel Online, and Google Sheets. These Connectors are designed to simplify data integration and maximize productivity for developers working with these platforms. From now on, it will be extremely easy to connect Python applications directly to Dynamics 365 Business Central, Excel Online, and Google Sheets, enabling accurate create, read, update, and delete operations on stored data.

As businesses realized the potential of artificial intelligence (AI), the race began to incorporate machine learning operations (MLOps) into their commercial strategies. But the integration of machine learning (ML) into the real world proved challenging, and the vast gap between development and deployment was made clear. In fact, research from Gartner tells us 85% of AI and ML fail to reach production.

In today’s digital-first world, organizations are constantly pressured to deliver faster, more reliable, and secure network services. Yet, with a staggering 65% of enterprise network activities still performed manually, the journey toward network automation is more critical than ever.

We’ve all been there: working a feature that spans multiple Git repositories, doing your best to coordinate pull requests and save the reviewers from “review-hell” as they try to piece together how everything fits across the repositories… When a change impacts more than one microservice, the coordination can become a mess of disconnected PRs spanning many repos, each dependent on the other to avoid breaking something critical.