In this keynote analysis from the Kong API Summit ‘25, theCUBE Research’s Paul Nashawaty breaks down how Kong is positioning Konnect as a unified platform for APIs, AI, agents and event streams – laying the groundwork for the agentic AI era. Nashawaty unpacks the major announcements around security, governance, monetization and unification designed to help enterprises securely scale AI- and agent-driven workloads. He highlights data points from theCUBE Research, including that 54% of production applications are already incorporating AI and that 68% of organizations prefer vendors that actively support open source – context for why Kong’s moves matter.

Nashawaty dives into the AI and MCP news: Konnect MCP support (including an MCP gateway, centralized authentication and enhanced observability/governance), an MCP-enabled Developer Portal with an AI assistant and AI integrations for Composer and Runner to execute intelligent endpoints securely. He spotlights the open-source Volcano SDK for building and testing MCP-powered AI agents. On security, he details “Kong Identity,” which unifies API and machine identity to secure the non-human internet with per-region auth servers, dynamic claim templates and developer self-service – advancing zero-trust for agentic systems. He also covers “KAi,” an intelligent Copilot for API and agentic operations inside Konnect. Additional updates include Konnect Service Catalog support for AWS (adding native AWS API Gateway integration to curb API sprawl across hybrid and multicloud) with governance scorecards for compliance, as well as Konnect Metering & Billing (building on the OpenMeter acquisition) to monetize APIs, AI models and data streams via real-time metering, flexible rate cards and automated billing.

The discussion closes with Kong Event Gateway extending Konnect to event-driven data systems by exposing Kafka streams as secure, discoverable APIs – unifying API and event governance under one control plane, visualizing clusters for multi-team efficiency and enabling self-service data access for AI agents and LLMs through the developer portal. Throughout, Nashawaty ties these moves to the broader shift theCUBE is tracking: intelligent gateways, AI-driven observability and event-based design converging so developers, platform teams and enterprise architects can build, secure and scale the next wave of connected applications.

In this Kong API Summit ‘25 segment, theCUBE Research’s Paul Nashawaty sits down with Ben van Leliveld, program manager of IT at DELTA Fiber, to unpack how the Netherlands-based service provider is building a future-proof architecture on Kong’s API gateway. Van Leliveld explains why DELTA Fiber standardized on a single gateway to reduce complexity, accelerate time to market and strengthen security across fast-growing B2C, B2B and wholesale lines of business. He details the company’s investment in a 10-gig fiber network and the imperative for IT systems – customer portals, self-service apps and supporting APIs – to match that speed and reliability. The conversation covers program execution, change management and training as “people-first” success factors, along with results from an RFP/POC that showed a 100% fit and local deployment expertise from Kong’s Netherlands team.

The discussion also highlights measurable outcomes and guardrails. Van Leliveld shares gains from a comparable project where DELTA Fiber’s second implementation ran ~300% faster and ~60% less costly by leaning on open standards and configuration over customization. He describes stricter SLAs for enterprise and wholesale customers, governance and audit requirements in regulated telco environments and a penetration-testing anecdote underscoring API security posture. Looking ahead, DELTA Fiber plans a company-wide consolidation to Kong in 2026, with interest from the network domain to align on a unified gateway – advancing the Summit’s broader themes around API platform thinking, intelligent gateways and building for scale with security, observability and time-to-value in mind.

In this conversation at the Kong API Summit 2025, theCUBE Research’s Paul Nashawaty sits down with Mark Beerens, senior product manager for cloud API management at ING Bank, to unpack ING’s shift from a monolithic core to a flexible, microservices-powered landscape. Beerens explains how ING adopted a gradual, no-“big bang” approach: carving out services, sharing learnings across teams and re-skilling engineers while balancing culture and technology. He details why a hybrid cloud design (control plane on Azure, data planes on-prem) and developer “paved roads” – including Kong plugins for authentication, authorization and observability – have been key to accelerating modernization without sacrificing governance and security. The discussion also touches on regulatory realities across regions and why security remains top-of-mind as a global financial institution.

The interview explores event highlights and what’s next: building a more mature API layer as a foundation for the agentic era, expanding analytics toward proactive, predictive insights (e.g., acting on API call spikes), evaluating service mesh options and continuing UI-to-pipeline automation for platform consumers. Nashawaty brings recent research context, such as 67% of respondents hiring generalists over specialists, to frame the talent and operating-model shifts required for containerization and microservices at scale. It’s a practical look at how platform teams can reduce friction, empower developers and evolve from API teams to API platforms using Kong’s ecosystem.

In this Kong API Summit ’25 interview, Jonas Persson, enterprise architect at SEB, joins theCUBE Research’s Paul Nashawaty to unpack SEB’s multi-year API transformation and what it takes to scale from “meetings and manual steps” to fully automated pipelines. Persson explains how open banking pressures pushed SEB to adopt Kong, shifting from a centralized ESB model to a decentralized, product-oriented approach where teams own their integrations and “don’t even see the gateway.” He details the move to a developer portal, governance that fits banking realities and how frictionless platform capabilities let developers focus on business outcomes rather than protocols and security plumbing. The discussion ties to the Summit’s broader theme: APIs, microservices and agentic infrastructure converging to power connected applications with intelligent gateways and event-driven design.

The conversation also explores modernization trade-offs across legacy and “new world” stacks – files, MQ, Kafka and ESB alongside API products – and why management sponsorship, skill evolution and platform abstraction matter. Persson shares how externalizing well-documented APIs accelerated new customer engagement (including in custody services), reflects on research showing teams skewing toward generalists, and outlines SEB’s AI trajectory: setting a foundation of reusable APIs to enable innovation, exploring intelligent agents and collaborating with NVIDIA as Sweden builds an AI center. Looking ahead to 2026, he expects AI to help drive the business and assist with legacy remediation – while Kong’s platform continues to remove friction as teams evolve from API “teams” to API “platforms.”

At Kong API Summit 2025, theCUBE Research’s Paul Nashawaty sits down with Prashant Gaonkar, who leads Cognizant’s global practice for enterprise orchestration, integration and supply chain. Gaonkar breaks down why process orchestration is the “glue” across 400–500 systems inside large enterprises – and how an integration “string engine” underpins end-to-end workflows for customer and internal processes. He reacts to Kong’s latest moves – expanding governance and observability with MCP and embracing hybrid AI, on-prem and multi-cloud neutrality – and explains why that flexibility helps customers modernize without ripping and replacing prior cloud and SaaS investments. The discussion also looks at Cognizant being named Partner of the Year and the trust built around Kong Konnect and Kong Gateway.

The conversation tackles the realities of AI in regulated industries: the need for explainability, traceability and audit trails (including a real underwriting example where an agent’s output failed quality checks). Gaonkar and Nashawaty explore API monetization’s growing pains – highlighting a recent bank throttling APIs after one small consumer drove ~80% of workload – and why business models, SLAs and “kill switches” for agents must evolve alongside chargebacks. They dig into human-in-the-loop accountability, multi-agent orchestration on Kong’s platform and why success ultimately comes from careful deployment, strong governance and joint customer case studies that prove acceleration, cost reduction and productivity gains.

In this Kong API Summit ‘25 segment, Ross Kukulinski, vice president of product management at Kong for Kong Konnect (Kong’s cloud-based API & AI connectivity platform), and Peter Marton, co-founder of OpenMeter and now product manager for Kong’s API & AI monetization product, join theCUBE Research’s Paul Nashawaty on the show floor. The discussion zeroes in on how AI is reshaping monetization models and why metering sits at the heart of billing for both APIs and AI workloads. Marton outlines the emerging mix of token-based AI monetization, traditional API monetization and a growing “success-based” pricing pillar. Kukulinski explains why Kong integrated OpenMeter’s open-source, universal metering stack into Konnect to support high-throughput event ingestion, flexible product catalogs and enterprise-grade billing that extends beyond API traffic.

The conversation dives into what’s new: second-level real-time metering for governance and customer transparency, data-driven product and rate cards that let product, platform or finance teams iterate on pricing without engineering tickets, and fine-grained entitlements to manage limits, overages and add-ons. They explore practical outcomes – internal show-back/chargeback, partner usage scenarios (e.g., loyalty programs) and downstream invoice flows that can feed Stripe, NetSuite or enterprise ERPs. Kukulinski also previews the Konnect Metering & Billing experience, including per-customer lifecycle views and export to data lakes (e.g., S3/Snowflake) for CFO-level reporting. The segment closes with timing: early-access signups for the Beta in November and a target of general availability by year-end.

In this interview from the Kong API Summit 2025, Augusto “Aghi” Marietti, co-founder and chief executive officer of Kong, joins theCUBE Research’s Paul Nashawaty to unpack a pivotal shift: moving from API calls to tokens in the agentic era. Marietti explains why API traffic and AI traffic are converging and how Kong is building a unified API and AI platform that standardizes connectivity across REST, MCP, events/streaming and more – complete with a true system of record so agents can discover, govern and operate across the enterprise. The discussion covers modernization paths that “meet customers where they are,” spanning on-prem mainframe wrappers, cloud-native on AWS, event streaming with Kafka and LLM/agent-driven teams – all under one platform for security, discoverability, observability and monetization.

The conversation dives into Summit announcements including the AI Gateway (LLM + MCP gateway for traffic management) and Volcano SDK, a TypeScript SDK to build MCP-powered agents in minutes (available at Volcano.dev and GitHub). Nashawaty highlights research signals such as widespread AI in production and skill-gap pressures, while Marietti details how Kong aligns with business leaders (architects, platform engineering, CIO/CTO) to reduce friction via a programmable management plane: “all roads lead to Konnect” for end-to-end visibility from an API call to a token. Additional topics include governance at the traffic layer, API product monetization (billing, metering, enforcement), free-to-start access to Konnect, and how GSIs help customers execute enterprise-scale transformation. Marietti also previews day-two focus on MCP integrations and an initial AI agent for APIs aimed at elevating platform teams.

In this Kong API Summit ’25 interview, theCUBE Research’s Paul Nashawaty sits down with Marco Palladino, co-founder and chief technology officer of Kong, to unpack how agentic AI is reshaping enterprise connectivity. Palladino explains why AI is “the next universal interface” and details Kong’s newest AI infrastructure advances: first-class support for the MCP protocol, new MCP Gateway capabilities in Kong AI Gateway 3.12 alongside existing LLM Gateway features and auto-generation of MCP servers from any RESTful API via a simple YAML file. He cites real-world progress from customers such as Prudential and addresses the well-known MIT finding that ~95% of organizations attempting AI transformation are failing – framing Kong’s mission to lower that number by abstracting routing, guardrails, security and governance into the platform.

The conversation explores how MCP has quickly become foundational for agentic AI (analogous to what REST has been for traditional APIs), why building an ecosystem of MCP tools is essential for capable agents and how Kong’s open-source Volcano SDK helps developers build MCP-native agents faster (while still relying on a governed, compliant, secure AI infrastructure layer). Viewers will hear how a unified control plane for LLM, MCP and traditional API traffic can reduce friction across heritage and modern environments, speed time-to-market and enable platform teams to scale observability, security and real-time connectivity for the agent era.

At the Kong API Summit 2025, Reza Shafii, senior vice president of product at Kong, joins theCUBE’s Paul Nashawaty to unpack how the convergence of AI and APIs is moving from hype to reality. Shafii shares details from the keynote, including KAi, Kong’s copilot for Konnect, and new MCP server support designed to simplify how agentic applications weave LLM context with secure, reliable and discoverable APIs. He explains how solid API foundations accelerate AI initiatives and breaks down how AI Composer and Runner can stitch multiple backend APIs into a single tool in minutes to deliver outcomes like customer sentiment.

The discussion also covers the debut of Kong Identity for machine-to-machine identity, why agent impersonation of “super users” is risky and how forthcoming regulations (with reporting beginning September 2026 and enforcement December 2027) raise the stakes. Shafii outlines the Konnect service catalog as a system of record behind the developer portal, and he and Nashawaty explore the push to “move fast without shooting yourself in the foot.” They cite notable dynamics from the field – such as organizations aiming for hourly code releases and developers spending only a third of their time writing code – as they look at how strong governance and platform control can free builders to innovate.

In this Kong API Summit ’25 interview, Rishi Singh, who leads the cloud-native practice at Capgemini Financial Services (Chicago), joins theCUBE Research’s Paul Nashawaty to unpack what it really takes to modernize heritage middleware and prepare for agentic AI. Singh explains why enterprises must first rationalize federated stacks across tools like MuleSoft, DataPower/IBM API Connect, WSO2 and Apigee, and build a well-structured API layer before layering AI on top. The discussion dives into emerging agent frameworks and protocols – MCP, A2A and LangChain agent models – with Singh cautioning that while MCP is open source and gaining traction, it’s not yet production-ready due to security gaps. He expects tighter data-privacy specifications to follow and stresses human-in-the-loop guardrails when orchestrating multi-agent workflows.

Nashawaty highlights Capgemini’s 200+ certified Kong professionals and its recent Kong API Partner Awards recognition, as the pair explore migration patterns to Kong, greenfield API-first development and tech-debt reduction – plus a fourth, fast-rising area: agentic workflows. Singh outlines practical KPIs that tie modernization to business outcomes (e.g., time-to-market and bi-weekly release cadence) and offers candid guidance for CIOs facing “use-it-or-lose-it” AI budgets: standardize services and middleware, start small with a focused use case (such as reducing premium-audit leakage in insurance), validate with a handful of agents, and only then scale. The conversation also touches on how Capgemini is collaborating with Kong around an AI gateway and LLM proxy approach while addressing MCP security considerations – aligning with the Summit’s broader themes of APIs, microservices and agentic infrastructure converging to shape the next era of connected applications.