Systems | Development | Analytics | API | Testing

As LLMs become central to AI-driven products like copilots and customer support chatbots, data science teams need to ensure the LLM performs well for the use case. The process of LLM evaluation ensures reliability, safety and performance in production AI systems. In this guide, we explore how to approach evaluations across development and production lifecycles, what frameworks to use, and how the integration between open-source MLRun and Evidently AI enables more scalable, structured testing.

It’s no longer theoretical – artificial intelligence has left research labs and entered production systems, generating a new breed of consumers – autonomous and intelligent agents. These autonomous AI agents are increasingly interacting with real-world APIs (application programming interfaces), which are sets of protocols and tools for building and integrating software applications.

Innovation isn't a buzzword at ThoughtSpot; it's the lifeblood of our business. It’s woven into our culture, driving us to constantly push boundaries and deliver exceptional value to our customers, our partners, and our team. Recently, I had the privilege of sitting down with a powerhouse panel of senior functional leaders for an “Innovation Spotlight” session. Our goal? To share how bold thinking translates into real-world impact across the many different facets of our organization.

In the last two parts of this series, we discussed How to Strengthen a ReAct AI Agent with Kong AI Gateway and How to Build a Single-LLM AI Agent with Kong AI Gateway and LangGraph. In this third and final part, we're going to evolve the AI Agent with multiple LLMs and Semantic Routing policies across them. In this blog post, we'll also explore new capabilities introduced in Kong AI Gateway 3.11 that support other GenAI infrastructures.

When we officially launched our Model Context Protocol (MCP) server integration on June 12, 2025, we weren't just adding another feature - we were fundamentally changing how data engineers interact with their tools. One month later, the transformation has exceeded our wildest expectations.

Have ever opened up a dashboard or report, and not known where to start exploring? Finding meaningful conclusions from a sea of charts and tables can be challenging and time-consuming. It's not always easy to see and understand the story your data is trying to tell, especially when you’re presented with a lot of information at once.

Fivetran brings its enterprise-grade replication capabilities to unstructured data, unifying data integration across all formats.

At the 2025 instalment of its WWDC event, Apple set out its long-term vision for how we think about platform strategy, AI integration and multi-device architecture. If you’re a CTO, staff engineer, or mobile lead, this wasn’t just a conference to watch, it was one to plan your entire roadmap around. What Apple revealed at this year’s WWDS will affect everything from your frontend stack to how your systems talk to hardware.

As artificial intelligence (AI) reshapes industries, it’s quietly revolutionizing the heart of IT: The data center. The explosive growth of AI workloads is driving up power usage, challenging cooling systems, and demanding a fundamental rethink of how we store and move data. In this new landscape, flash storage stands out – delivering the performance, efficiency, and scalability that AI needs to truly accelerate.