Webinar

Mathias Bilmann at Netlify

With

The New Frontier in AI Development: Why Agent Experience Matters

21 May 2025

Agent experience (AX) is the holistic experience AI agents will have as the user of a product or platform. AI agents are reshaping software development, but today’s developer tools and platforms weren't built for them. For AI agents to truly be effective, we need to craft our products with AX in mind. How do we design platforms, APIs, and workflows that help AI agents operate effectively? While DX optimized tools for humans, AX challenges us to rethink systems for AI-first development.

Defining Agent Experience (AX): The Next Evolution in Software

In his keynote "The New Frontier in AI Development: Why Agent Experience (AX) Matters," Mathias Biilmann, CEO and co-founder of Netlify, framed AX as a crucial, emerging discipline in the world of AI-powered software. Biilmann began by discussing the history of user experience (UX), referencing Don Norman and Jakob Nielsen’s seminal work. He explained how UX revolutionized software by emphasizing the importance of the human experience beyond just a checklist of features. Drawing a parallel, Biilmann argued that the industry now stands at a similar inflection point with agent experience—where the primary "users" of products are increasingly AI agents, not just humans.

Netlify's Path to AX: From Developer Focus to Agent-Centric Design

Biilmann described how the concept of AX originated in a brainstorming session with a principal engineer at Netlify, as they noticed the growing importance of coding agents integrating with Netlify’s deployment platform. They realized that optimizing features for agents, not just humans, required its own term and thus coined "AX." Biilmann asserted, “AX is the holistic experience an AI agent will have as the user of a product or platform,” marking a shift towards products that cater to both human and agent workflows.

Reflecting on Netlify’s trajectory, Biilmann shared how developer experience (DX) became a differentiator for the company’s success as they moved from traditional web architectures to empowering front-end developers. He noted that developer experience involved not just APIs and features, but frictionless onboarding, great documentation, and a native feel in developer tools. Now, Biilmann argued, AX is becoming just as central, stating, “Anyone that is not building a great agent experience is not going to be able to build a great developer experience either.”

AX in Action: Real-World Implementations and Industry Impact

Describing practical patterns, Biilmann highlighted Netlify’s early move towards AX when, in May 2023, OpenAI launched the GPT Store. Netlify built a custom GPT agent that allowed instant site deployment without friction—no signup or authorization, just an agent operating on the user’s behalf. This seamless workflow drove explosive growth, as Biilmann reported “more than 1,000 sites created every single day just from the GPT ecosystem” at that time, scaling soon after to “about 10,000 sites created directly from coding agents every single day.”

He detailed collaborations with partners like Bolt and WindServe, enabling agent-driven deploy-and-claim workflows: “completely frictionless onboarding for an agent doing work on behalf of a user.” Biilmann described how these patterns lower the adoption barrier and are quickly becoming industry standards, citing similar flows adopted by Prisma and Niantipi.

Technical Imperatives for AX: New Standards and Architectural Shifts

On the technical side, Biilmann emphasized the need for standards and new protocols—central among them, Model Context Protocol (MCP). He shared examples showing that best practices for AX often differ from those for DX; for instance, “LLMs work much better with XML than with JSON,” contrary to developer preferences. He also referenced Marcelo Torero at Bricks, who “re-architected their front-end to be more LLM friendly,” highlighting the need to rethink architectures for agents, not just humans.

Documentation and API context emerged as critical, as agents may use outdated information or tutorials from across the web. Patterns for injecting “simple versions of docs into LLM context become incredibly important,” Biilmann added, observing that “good DX might not be good AX.”

The Road Ahead: Co-Creating an Agent-First Future

Addressing concerns about developers’ roles in an agent-driven world, Biilmann was optimistic. He predicted that AX would “lower the barrier” for future developers, enabling “the next 100 million” to participate, even if much of their work happens indirectly through agents.

Looking toward the future, Biilmann advocated for industry-wide standards, inviting collaboration through community platforms like agent-experience.ax. He concluded by emphasizing that adopting AX means building a whole new discipline—one requiring empathy for both human and agent users and open-mindedness as traditional development workflows evolve.

Ultimately, Biilmann’s message was clear: for companies building the next generation of products powered by AI, prioritizing agent experience is not just an option, but a necessity to remain competitive and relevant in an agent-augmented software landscape.

Subscribe to our podcasts here

Welcome to the AI Native Dev Podcast, hosted by Guy Podjarny and Simon Maple. If you're a developer or dev leader, join us as we explore and help shape the future of software development in the AI era.

THE WEEKLY DIGEST

Subscribe

Sign up to be notified when we post.

Subscribe

JOIN US ON

Discord

Come and join the discussion.

Join