Systems | Development | Analytics | API | Testing

As we explored in our previous post, SAP’s ecosystem makes it costly and complex to centralize data, especially when organizations need to blend SAP and non-SAP sources for unified analytics. Between restrictive licensing, limited compatibility, and constrained export options, SAP’s architecture often creates more roadblocks than results.

The era of generative AI is upon us, and large language models (LLMs) like Claude are demonstrating incredible potential to revolutionize how businesses operate and interact with customers. However, to truly unlock this potential, AI needs secure and standardized access to the wealth of information and services locked within enterprise systems. This is where standards such as Model Context Protocol (MCP) come in, offering a powerful way to make enterprise resources AI-consumable.

Devoxx Poland 2025 felt different. Not because of revolutionary new frameworks or another "this changes everything" moment, but because of what didn't happen. The conference had an unusual dose of pragmatism, skepticism, and – dare we say it – common sense. Maybe it's because developers are asking the right questions: "Does this solve a problem?" and "What happens when this inevitably breaks?" Here's what emerged from the sessions we watched, and the people we spoke to.

When Java operations hit their time limits, they throw java.util.concurrent.TimeoutException. This checked exception appears in scenarios ranging from thread synchronization barriers that never complete to Selenium tests waiting for elements that never appear. The challenge isn't just catching these exceptions—it's designing your code to handle timing constraints intelligently.

In a new Forbes Insights report, 500 C-suite tech leaders share how they’re navigating data complexity, scaling AI, and prioritizing governance — and why modern integration is now mission-critical.

How can SAP customers break free from ecosystem constraints to securely integrate SAP data with the rest of their stack?

In case you missed it, we recently made a big announcement around beta GCP support for Kong’s Dedicated Cloud Gateways (DCGWs). There’s a lot of good stuff in there, but TL;DR DCGWs now support all three of the major cloud service providers (CSPs): AWS, Azure, and GCP at a 99.95% SLA with support for over 25 regions around the globe. Being the first API management vendor to support managed gateway deployments with all three CSPs has a lot of folks excited, for obvious reasons.

Enterprise adoption of large language models (LLMs) is surging. According to Gartner, more than 80% of enterprises will have deployed generative AI (GenAI) applications or used GenAI APIs by 2026, up from just 5% in 2023. That stark increase paints a telling picture: LLMs have evolved from a fringe technology to a cornerstone of business development and productivity. But as with any new technology, competition is fierce.

Let’s take a moment to think about how far we’ve come. Remember when opening a bank account meant sitting at a branch, filling out stacks of paperwork? Or when transferring money meant writing a cheque and hoping it cleared in a few days? Now, you can do all of that and so much more, with just a few taps on your phone. So, what changed? And more importantly, what’s driving this massive shift in how we bank today? The answer lies in one word: AI in Banking.

You’ve probably heard of the concept of ‘Frontend decides, backend delivers’ in app development. On the off-chance that you haven’t, it means that the frontend defines the data it needs, and the backend acts on this instruction. This makes the data-fetching process more efficient, simplifies the error handling process and frees us, the devs, from the need to constantly make backend changes. The GraphQL query language for APIs, developed by Facebook, is a vital tool in this regard.