AI agents are transforming the way we build — and even how we think of ourselves as software developers. Both Spotify and Anthropic have been exploring what this shift means in practice, and on March 30, we got together at Spotify’s London HQ to share what we’re learning.
In our fireside chat “Let’s Talk Agentic Development”, Spotify’s chief architect, Niklas Gustavsson, sat down with Anthropic’s David Soria Parra (co-creator of MCP) and Christian Ryan (applied AI lead) for a wide-ranging discussion on what agentic-first development actually looks like at scale — from the tools and infrastructure that make it work, to the organizational shifts it demands. (View Highlight)
Context and control: Foundations for AI at scale
How to standardize your ecosystem and orchestrate Claude across thousands of repos. Both teams share what’s actually working — and where the gaps still are — when it comes to giving agents the right context at enterprise scale. (View Highlight)
Humans vs. agents: Testing, reviews, and governance
What changes when agents can ship code faster than humans can review it? The panel gets into the new bottlenecks — and the hard questions about accountability — that come with scaling agent-generated output.
“It doesn’t really matter who generated what or what was behind it. If it’s an agent or a human, it’s very much outcome-based, and you also want to have someone who’s accountable for the outcome.” — Christian Ryan, Anthropic (View Highlight)
Toward the end of the conversation, David points out that 2025 has been about the creation of code, but the next frontier is agents taking on the full software lifecycle — maintenance, deletion, and the kind of work nobody wants to do but everyone needs.
Backstage is evolving from a human-facing developer portal into an agent-first platform, with MCP connections replacing manual workflows. (View Highlight)
