Systems | Development | Analytics | API | Testing

Kong AI Gateway 2.0 can be installed on Kubernetes in about five minutes using Kong Operator, Helm, and a self-managed control plane in Konnect. This walkthrough starts with a two-node cluster, creates a new AI Gateway in the Konnect UI, and installs Kong Operator with Helm. It then generates a personal access token, applies the API auth secret, and deploys the data plane through custom resources. You’ll see how to configure replicas, image, and load balancer ports in the YAML. This is the first video in the Kong in 5 series for Kong customers and users.

See how an airline company can use Confluent to build a Streamhouse that powers a real-time flight operations dashboard and AI agents to rebook passengers using current flight and hotel data. Explore Lightning Tables, Confluent Intelligence (Streaming Agents), and Tableflow.

See how Lightning Tables in Confluent Cloud serves the current state of of the business as a fully managed read-only layer—so AI agents and applications can query fresh context at low latency and high concurrency, without ETL, separate databases, or caches. The demo shows how to enable Lightning Tables on Kafka topics for instant querying.

AI is changing what businesses need from their data. Applications and AI agents need access to current, governed business context as events unfold, not after another refresh cycle. Enter the Streamhouse: an architectural approach designed to continuously capture and govern live data, keeping business context current across applications, analytics and AI. Built on three core principles—always current, mission-critical and decentralized—a Streamhouse provides the foundation for making real-time data usable across the enterprise.

A $6,000 laptop could change the math on a $60,000 AI token bill. That fact makes Patrick Moorhead question how businesses calculate the cost of AI. For suitable workloads, running an open model locally could make an expensive-looking hardware purchase a worthwhile investment. His focus is the cost of producing a reliable result.

While it may seem harmless to give your AI a name, or a gender, you could be inadvertently opening up a security risk in the process. Dr. Peter Garraghan has spent years studying how attackers exploit AI systems, and some of his findings are surprisingly human. AI models can be manipulated through language, using techniques that resemble the social engineering tactics cybersecurity professionals have dealt with for decades.