Kong

What happens when a “simple” account update touches multiple systems? Build a governed, self-service account update workflow with Docusign + @kong Konnect, from API security to backend integration. ⚒️ Explore the guide. docusign.com/blog/developer…

MCP 2026-07-28 is live and it's the largest update to the protocol since launch. MCP is now stateless, making it easier to deploy and scale remote servers. claude.com/blog/bringing-…

What does it actually take to run AI in production? Join us at API + AI Summit 2026 to learn directly from companies like LinkedIn, Spotify, Southwest Airlines, AWS, Dell, Pandora, and more! ✨ They'll be sharing practical lessons on AI governance, agentic architectures, infrastructure, and operating AI. Check out the 7 can't-miss sessions here: bit.ly/4fL15b4

We found something crazy in our AI gateway data governing model and mcp traffic for hundreds of enterprises. One of our customers is running 602 different AI models in production right now. Not a typo. Six hundred and two distinct model identifiers, all proxied through a single AI Gateway. 20% of organizations on our platform run six or more models in production at the same time. 62% run two or more. Multi-model is the default now. Cost-sensitive tasks go to smaller, faster models. Hard reasoning goes to the frontier. Different teams pick different providers, and every time a new model is released, we can see the portfolio change again.

The new Konnect Dev Portal Toolkit from @kong brings live previews of your Konnect Developer Portal right into VS Code✨ Pair it with the Konnect MCP and let AI agents author, brand, validate, and visually verify Dev Portal content on their own. konghq.com/blog/engineeri…

🐤🤝 🦍 We just expanded our @kong partnership to cover Kong AI Gateway too. Every agent, MCP server, and LLM service now gets auto-discovered and monitored, plus real-time protection against prompt injection and data leaks. Learn more: bit.ly/4pDBfds

This week marks a special milestone for Kong as we opened our newest office in Milan - a return to the city where our co-founders first started. 🍝 It's one of several milestones driving our momentum across EMEA: 🏢 New office in Paris + expanded offices in London & Budapest 📈 72% YoY growth in our EMEA customer base 👥 71% growth in our regional team over the past two years You can learn more about Kong's growth across EMEA here: konghq.com/company/press-…

The must-see premiere of the fall isn't on the big screen. 🎬 It's API + AI Summit 2026: Where the engineers, architects, and AI leaders building the future of AI are coming together. 🎟️ Get your tickets today: bit.ly/4yCv8u8

Who will be at Ai4 in a couple of weeks? 🙋 Stop by booth #1021 to meet the team, get a demo, and talk about how Kong enables builders to fully focus on building agents, and helps you control costs for all AI resource consumption and monetize agent, MCP and API access. Let us know if we'll see you there!

Cool chatting on AI governance @MTSlive. Among the largest enterprises we work with, kong gateways are becoming both the observability and chokepoints of AI as agents, models and APIs connect on the token-path. The world is moving from tokenmaxxing to token cost management. The best way to unlock AI is to govern it at the traffic layer.

Kong CEO Aghi Marietti on the air traffic control analogy he uses to explain why every enterprise deploying AI is one bad landing from disaster: "Think about an AI traffic control tower at an airport. Those are the busiest airports, they can have a landing every 30 seconds." "The only reason there are so many planes from so many runways landing and taking off is because there's a central control tower that manages and sequences and securely allows flights to come in and out." "The same thing on AI in an enterprise: you have an AI control tower where it can sequence, secure, and make safe all the AI coming in and out." "In an enterprise, it's Wild West as of today. What we provide is digital railroading, guardrails to agents." "The agent can still have intelligence to make its own decisions, but it won't screw up a 40-year-old company in public, because you can put guardrails on what agents, APIs, and LLMs can do or cannot do." @sonicaghi @kong

Happy National AI Day! P.S.: Meet Kong's new official mascot, Karl. Karl, meet everybody. 👋

Scary thought you might be having: Right now, your token usage is the LOWEST it will ever be. Which is especially terrifying if you're a CFO that's only seen a 1:1 relationship between token usage and costs. It doesn't have to be like that. You can actually use more tokens and spend LESS if you build the right infra. Semantic routing might be the most effective way to cut token spend, but not enough orgs are prioritizing it. With an AI gateway, you get a decision layer that maps the intent of requests to the most appropriate model for each job. You should not be spending frontier model prices on something like log summarization. Semantic routing fixes that. Simple, high-volume queries get sent to fast and cheap models automatically. Complex reasoning goes to the frontier. If you do it right, your token usage keeps climbing but your cost curve won't follow it.

As API footprints grow in scale and complexity, the gap between what platform teams need and what it takes to operate their infrastructure efficiently continues to grow. API Gateway 3.15 closes that gap. 🎉 What's new in 3.15? Plugin cloning, plugin streaming, security enhancements, conditional policy execution, distroless images + portal application traffic governance. Learn more: bit.ly/4aWSKiL

New coat of paint for our @kong docs! 🎨

Why do I always say there's "no AI without APIs"? If you read the phrase literally, you'd actually be right to disagree with me. Technically, a model doesn't need an API to run local tasks on a GPU, but APIs are the connective tissue of AI. The nervous system carrying signals in and out to add context and meaning. Without APIs your model is just a brain in a box. APIs are what move artificial intelligence. Storage holds information. Compute transforms it. Connectivity (APIs, MCP servers, the traffic layer) is what moves it. The third leg of the stool, and the thing the industry isn't talking about enough. No interface, no action. No APIs, no functional AI.

AI runs on APIs. And moving AI into production requires governance, observability, and security. 🎙️ Our VP of Solutions Engineering, Felderi Santiago, joined @gigaom's Business Disruptions in Tech podcast to discuss building enterprise-ready AI and agentic systems with a strong API foundation. Watch the full episode here: bit.ly/4w3EJbD

Every new hire at Kong has to build an API in their onboarding process. We think every person working here needs to understand the invisible foundation of modern software (APIs). Doesn't matter if you're in sales, finance, or engineering. Some of our recent "API of the month" winners: Sean Keever: "Lord of the Pings" A serverless API gateway forged in onboarding to route every request, bind the backends, and keep Middle-earth’s services answering one ping at a time. Kratagya Parashar: "The Kong Father" An API that made its services an offer they couldn't refuse. Guido Salvatore Spina: "The Lasagna Factory" Because even Nonna's lasagna deserves enterprise-grade orchestration.

Hot take: Your Vectordb is a terrible choice for being your agent's only memory system. 🔥 Vectors are great for retrieval. They're terrible at ordering and replay. Agents need both semantic memory and durable history. That's why Kafka matters. buff.ly/17eJOMn

The incredible @Kong team just opened our beautiful Paris office! 🇫🇷 After years of hyper-growth in France and around europe we decided to double down with a local hub to bring together Customer Success, Sales, Professional Services, and many other functions closer to our growing customers base. Pas d'IA sans API... EN AVANT! 🦍 PS: of course the soccer match could not be missed. ⚽