The New Stack Podcast

AI applications are evolving beyond chatbots into more complex and transformative solutions, according to Marco Palladino, CTO and co-founder of Kong. In a recent episode of The New Stack Makers, he discussed the rise of AI agents, which act as "virtual employees" to enhance organizational efficiency. For instance, AI can now function as a product manager for APIs—analyzing documentation, detecting inaccuracies, and making corrections.

However, reliance on AI agents brings security risks, such as data leakage and governance challenges. Organizations need observability and safeguards, but developers often resist implementing these requirements manually. As GenAI adoption matures, teams seek ways to accelerate development without rebuilding security measures repeatedly.

To address these challenges, Kong introduced AI Gateway, an open-source plugin for its API Gateway. AI Gateway supports multiple AI models across providers like AWS, Microsoft, and Google, offering developers a universal API to integrate AI securely and efficiently. It also features automated retrieval-augmented generation (RAG) pipelines to minimize hallucinations.

Palladino emphasized the need for consistent security in AI infrastructure, ensuring developers can focus on innovation while leveraging built-in protections.

Learn more from The New Stack about Kong’s AI Gateway

Kong: New ‘AI-Infused’ Features for API Management, Dev Tools

From Zero to a Terraform Provider for Kong in 120 Hours

Join our community of newsletter subscribers to stay on top of the news and at the top of your game.

Platform engineering was meant to ease the burdens of Devs and Ops by reducing cognitive load and repetitive tasks. However, building internal development platforms (IDPs) has proven challenging. Despite this, Gartner predicts that by 2026, 80% of software engineering organizations will have a platform team.

In a recent New Stack Makers episode, Mallory Haigh of Humanitec and Nathen Harvey of Google discussed the current state and future of platform engineering. Haigh emphasized that many organizations rush to build IDPs without understanding why they need them, leading to ineffective implementations. She noted that platform engineering is 10% technical and 90% cultural change, requiring deep introspection and strategic planning.

AI-driven automation, particularly agentic AI, is expected to shape platform engineering’s future. Haigh highlighted how AI can enhance platform orchestration and optimize GPU resource management. Harvey compared platform engineering to generative AI—both aim to reduce toil and improve efficiency. As AI adoption grows, platform teams must ensure their infrastructure supports these advancements.

Learn more from The New Stack about platform engineering:  

Platform Engineering on the Brink: Breakthrough or Bust?

Platform Engineers Must Have Strong Opinions

The Missing Piece in Platform Engineering: Recognizing Producers

Join our community of newsletter subscribers to stay on top of the news and at the top of your game.

AI agents are set to transform software development, but software itself isn’t going anywhere—despite the dramatic predictions. On this episode of The New Stack Makers, Mark Hinkle, CEO and Founder of Peripety Labs, discusses how AI agents relate to serverless technologies, infrastructure-as-code (IaC), and configuration management. 

Hinkle envisions AI agents as “dumb robots” handling tasks like querying APIs and exchanging data, while the real intelligence remains in large language models (LLMs). These agents, likely implemented as serverless functions in Python or JavaScript, will automate software development processes dynamically. LLMs, leveraging vast amounts of open-source code, will enable AI agents to generate bespoke, task-specific tools on the fly—unlike traditional cloud tools from HashiCorp or configuration management tools like Chef and Puppet. 

As AI-generated tooling becomes more prevalent, managing and optimizing these agents will require strong observability and evaluation practices. According to Hinkle, this shift marks the future of software, where AI agents dynamically create, call, and manage tools for CI/CD, monitoring, and beyond. Check out the full episode for more insights. 

Learn more from The New Stack about emerging trends in AI agents: 

Lessons From Kubernetes and the Cloud Should Steer the AI Revolution

AI Agents: Why Workflows Are the LLM Use Case to Watch 

Join our community of newsletter subscribers to stay on top of the news and at the top of your game. 

The transition from SaaS to Services as Software with AI agents is underway, necessitating new orchestration methods similar to Kubernetes for containers. AI agents will require resource allocation, workflow management, and scalable infrastructure as they evolve. Two key trends are driving this shift: 

Data Evolution – From spreadsheets to AI agents, data has progressed through relational databases, big data, predictive analytics, and generative AI. 

Computing Evolution – Starting from mainframes, the journey has moved through desktops, client servers, web/mobile, SaaS, and now agentic workflows. 

Janakiram MSV, an analyst, notes on this episode of The New Stack Makers that SaaS depends on data—without it, platforms like Salesforce and SAP lack value. As data becomes more actionable and compute more agentic, a new paradigm emerges: Services as Software. AI agents will automate tasks previously requiring human intervention, like emails and sales follow-ups. However, orchestrating them will be complex, akin to Kubernetes managing containers. Unlike deterministic containers, AI agents depend on dynamic, trained data, posing new enterprise challenges in memory management and infrastructure. 

Learn more from The New Stack about evolution to AI agents:

 How AI Agents Are Starting To Automate the Enterprise 

Can You Trust AI To Be Your Data Analyst? 

Agentic AI is the New Web App, and Your AI Strategy Must Evolve 

Join our community of newsletter subscribers to stay on top of the news and at the top of your game. 

Amazon Q Developer is streamlining the software development lifecycle by integrating AI-powered tools into AWS. In an interview at AWS in Seattle, Srini Iragavarapu, director of generative AI Applications and Developer Experiences at AWS, discussed how Amazon Q Developer enhances the developer experience. Initially focused on inline code completions, Amazon Q Developer evolved by incorporating generative AI models like Amazon Nova and Anthropic models, improving recommendations and accelerating development. British Telecom reported a 37% acceptance rate for AI-generated code.

Beyond code completion, Amazon Q Developer enables developers to interact with Q for code reviews, test generation, and migrations. AWS also developed agentic frameworks to automate undifferentiated tasks, such as upgrading Java versions. Iragavarapu noted that internally, AWS used Q Developer to migrate 30,000 production applications, saving $260 million annually. The platform offers code generation, testing suites, RAG capabilities, and access to AWS custom chips, further flattening the SDLC by automating routine work. Listen to The New Stack Makers for the full discussion.

Learn more from The New Stack about Amazon Q Developer: 

Amazon Q Developer Now Handles Your Entire Code Pipeline 

Amazon Q Apps: AI-Powered Development for All 

Amazon Revamps Developer AI With Code Conversion, Security 

Join our community of newsletter subscribers to stay on top of the news and at the top of your game. 

Maya Kaczorowski noticed that AI identity and AI agent identity concerns were emerging from outside the security industry, rather than from CISOs and security leaders. She concluded that OAuth, the open standard for authentication, already serves the purpose of granting access without exposing passwords. 

Kaczorowski, a respected technologist and founder of Oblique, a startup focused on self-serve access controls, recently wrote about OAuth and AI agents and shared her insights on this episode of The New Stack Makers. She noted that developers see AI agents as extensions of themselves, granting them limited access to data and capabilities—precisely what OAuth is designed to handle. 

The challenges with AI agent identity are vast, involving different approaches to authentication, such as those explored by companies like AuthZed. While existing authorization models like RBAC or ABAC may still apply, the real challenge lies in scale. The exponential growth of AI-related entities—from users to LLMs—could mean even small organizations manage hundreds of thousands of agents. Future solutions must accommodate this massive scale efficiently. 

For the full discussion, check out The New Stack Makers interview with Kaczorowski. 

Learn more from The New Stack about OAuth requirements for AI Agents: 

OAuth 2.0: A Standard in Name Only? 

AI Agents Are Redefining the Future of Identity and Access Management

Join our community of newsletter subscribers to stay on top of the news and at the top of your game. 

The rise of the World Wide Web enabled developers to build tools and platforms on top of it. Similarly, the advent of large language models (LLMs) allows for creating new AI-driven tools, such as autonomous agents that interact with LLMs, execute tasks, and make decisions. However, verifying these decisions is crucial, and critical reasoning may be a solution, according to Yam Marcovitz, tech lead at Parlant.io and CEO of emcie.co.

Marcovitz likens LLM development to the evolution of programming languages, from punch cards to modern languages like Python. Early LLMs started with small transformer models, leading to systems like BERT and GPT-3. Now, instead of mere text auto-completion, models are evolving to enable better reasoning and complex instructions.

Parlant uses "attentive reasoning queries (ARQs)" to maintain consistency in AI responses, ensuring near-perfect accuracy. Their approach balances structure and flexibility, preventing models from operating entirely autonomously. Ultimately, Marcovitz argues that subjectivity in human interpretation extends to LLMs, making perfect objectivity unrealistic.

Learn more from The New Stack about the evolution of LLMs: 

AI Alignment in Practice: What It Means and How to Get It 

Agentic AI: The Next Frontier of AI Power 

Make the Most of AI Agents: Tips and Tricks for Developers 

Join our community of newsletter subscribers to stay on top of the news and at the top of your game. 

Adam Jacob, CEO of System Initiative, discusses a shift in infrastructure automation—moving from writing code to building models that enable rapid simulations and collaboration. In The New Stack Makers, he compares this approach to Formula One racing, where teams use high-fidelity models to simulate race conditions, optimizing performance before hitting the track.

System Initiative applies this concept to enterprise automation, creating a model that understands how infrastructure components interact. This enables fast, multiplayer feedback loops, simplifying complex tasks while enhancing collaboration. Engineers can extend the system by writing small, reactive JavaScript functions that automate processes, such as transforming existing architectures into new ones. The platform visually represents these transformations, making automation more intuitive and efficient.

By leveraging models instead of traditional code-based infrastructure management, System Initiative enhances agility, reduces complexity, and improves DevOps collaboration. To explore how this ties into the concept of the digital twin, listen to the fullNew Stack Makers episode.

Learn more from The New Stack about System Initiative:

Beyond Infrastructure as Code: System Initiative Goes Live

How System Initiative Treats AWS Components as Digital Twins

System Initiative Code Now Open Source

Join our community of newsletter subscribers to stay on top of the news and at the top of your game. 

Morgan McLean, co-founder of OpenTelemetry and senior director of product management at Splunk, has long tackled the challenges of observability in large-scale systems. In a conversation with Alex Williams onThe New Stack Makers, McLean reflected on his early frustrations debugging high-scale services and the need for better observability tools.

OpenTelemetry, formed in 2019 from OpenTracing and OpenCensus, has since become a key part of modern observability strategies. As a Cloud Native Computing Foundation (CNCF) incubating project, it’s the second most active open source project after Kubernetes, with over 1,200 developers contributing monthly. McLean highlighted OpenTelemetry’s role in solving scaling challenges, particularly in Kubernetes environments, by standardizing distributed tracing, application metrics, and data extraction.

Looking ahead, profiling is set to become the fourth major observability signal alongside logs, tracing, and metrics, with general availability expected in 2025. McLean emphasized ongoing improvements, including automation and ease of adoption, predicting even faster OpenTelemetry adoption as friction points are resolved.

Learn more from The New Stack about the latest trends in Open Telemetry:

What Is OpenTelemetry? The Ultimate Guide

Observability in 2025: OpenTelemetry and AI to Fill In Gaps

Honeycomb.io’s Austin Parker: OpenTelemetry In-Depth

Join our community of newsletter subscribers to stay on top of the news and at the top of your game. 

Observability is expensive because traditional tools weren’t designed for the complexity and scale of modern cloud-native systems, explains Christine Yen, CEO of Honeycomb.io. Logging tools, while flexible, were optimized for manual, human-scale data reading. This approach struggles with the massive scale of today’s software, making logging slow and resource-intensive. Monitoring tools, with their dashboards and metrics, prioritized speed over flexibility, which doesn’t align with the dynamic nature of containerized microservices. Similarly, traditional APM tools relied on “magical” setups tailored for consistent application environments like Rails, but they falter in modern polyglot infrastructures with diverse frameworks.

Additionally, observability costs are rising due to evolving demands from DevOps, platform engineering, and site reliability engineering (SRE). Practices like service-level objectives (SLOs) emphasize end-user experience, pushing teams to track meaningful metrics. However, outdated observability tools often hinder this, forcing teams to cut back on crucial data. Yen highlights the potential of AI and innovations like OpenTelemetry to address these challenges.

Learn more from The New Stack about the latest trends in observability:

Honeycomb.io’s Austin Parker: OpenTelemetry In-Depth

Observability in 2025: OpenTelemetry and AI to Fill In Gaps

Observability and AI: New Connections at KubeCon

Join our community of newsletter subscribers to stay on top of the news and at the top of your game.