NerdOut@Spotify

At Spotify, we're doubling down on background coding agents. The goal is to scale code migrations and boost developer productivity beyond our existing automation system. Built on top of Fleet Management and supercharged by Claude Code, these agents make their own code changes and pull requests. What could possibly go wrong?
Engineers Jo Kelly-Fenton, Aleksander Mitic, and Max Charas join host Dave Zolotusky to unpack how we’ve merged thousands of AI-generated PRs while keeping codebases consistent and secure. They’ll dig into what it takes to let agents touch production code at scale: sandboxing and security, handling context windows, wiring into CI, and building verification loops that still trust humans more than agents.
You’ll hear real-world failure modes (like agents “fixing” tests by deleting them), how using an “LLM as a judge” flags sketchy diffs before they land, and how engineers and PMs are now triggering agents directly from Slack for migrations, documentation fixes, and feature flag rollouts. 
It’s a candid look at what increasing reliance on AI-based tooling means for the future of engineering — and for safely shipping code at speed.
Learn more about our journey with background coding agents:
1500+ PRs Later: Spotify’s Journey with Our Background Coding Agent (Part 1)

Background Coding Agents: Context Engineering (Part 2)

Background Coding Agents: Predictable Results Through Strong Feedback Loops (Part 3)

AI Engineer Paris 2025: Rewriting all of Spotify's code base, all the time

Read what else we’re nerding out about on the Spotify Engineering Blog: engineering.atspotify.com
You should follow us on Twitter @SpotifyEng, LinkedIn, and YouTube!

In part two of ARM-ing ourselves with Google Axion, Dave is joined by Max Charas, a senior staff engineer at Spotify who’s been leading much of the migration work on our Google Axion journey, for a deep dive into the validation and technical side of our move to Google’s new ARM-based processors.
Dave and Max unpack what went smoothly, what got bumpy, and what you only discover when you try to run thousands of services on brand-new silicon. Expect the usual nerdiness — performance gains, per-core scaling, bin-packing puzzles — as well as look at the less glamorous bits like discount models, capacity planning, and the logistical reality of working with real hardware. Because as it turns out, the hardest parts weren’t always technical. It was the massive coordination effort behind them that became a logistical puzzle on a global scale.
This episode is a practical look at how a large-scale system evolves when its foundation shifts — and how much just works when you’ve built solid abstractions.
Learn more about Google Axion Processors: 
Leading processors custom-built for cloud workloads

Introducing Google Axion Processors, our new Arm-based CPUs

C4A VMs now GA: Our first custom Arm-based Axion CPU

Axion in action: Unlock price-performance and efficiency with Google Axion VMs

The Impact of Google Cloud’s Arm-Based Axion Chip: Explained

Benchmarks Of Google's Axion Arm-based CPU: Competitive Performance & Compelling Value

Read what else we’re nerding out about on the Spotify Engineering Blog: engineering.atspotify.com
You should follow us on Twitter @SpotifyEng, LinkedIn, and YouTube!

Spotify is in the middle of a big shift in modern computing: moving from x86 processors to Google’s new ARM-based Axion chips. This isn’t just a hardware swap — it’s a disruptive change with ripple effects across performance, efficiency, and sustainability.
In this episode, we sit down with Mo Farhat, Group Product Manager for Google Compute Engine, to unpack what makes ARM so game-changing in the data center. From the history of ARM’s rise, to the challenges of designing chips for hyperscale workloads, to why this transition matters for the future of cloud computing, Mo offers a look inside how Axion was built and why now is the tipping point for ARM.
If you’ve ever wondered what really powers the services you use every day — and why Spotify is betting on Axion — this is the episode to listen to.
Learn more about Google Axion Processors: 
Leading processors custom-built for cloud workloads

Introducing Google Axion Processors, our new Arm-based CPUs

C4A VMs now GA: Our first custom Arm-based Axion CPU

Axion in action: Unlock price-performance and efficiency with Google Axion VMs

The Impact of Google Cloud’s Arm-Based Axion Chip: Explained

Read what else we’re nerding out about on the Spotify Engineering Blog: engineering.atspotify.com

You should follow us on Twitter @SpotifyEng, LinkedIn, and YouTube!

Hear about the journey behind building Spotify’s AI knowledge assistant, aka AiKA — from hack projects to internal chatbot, to enterprise developer tool.
While general purpose chatbots are being used everywhere, they often fall short when it comes to navigating company-specific information — like which internal policy applies to your team or where to find the doc someone shared in Slack three months ago. That’s why we built AiKA. Using retrieval-augmented generation (RAG), AiKA taps into Spotify’s internal knowledge sources to provide employees with context-aware answers right when they need them.
Host and principal engineer Dave Zolotusky is joined by senior engineers Majd Salman and Jofre Mateu to discuss AiKA’s evolution from a bunch of hack week experiments to a unified chatbot platform now used weekly by 25% of Spotify’s employees and 87% of our developers. They share why RAG is the right approach for making an internal chatbot accurate and fast, how AiKA has cut the time it takes to resolve internal support requests by 47%, and how we’re extending AiKA’s agentic capabilities with MCP.
🤖 Want to see how AiKA can supercharge knowledge sharing where you work? Sign up to try Spotify Portal for Backstage at: https://backstage.spotify.com/try-portal/ 
Learn more about Spotify’s AI knowledge assistant:
Our KubeCon talk: Leveraging Internal Knowledge: Building AiKA at Spotify
The New Stack: Introducing AiKA: Backstage Portal AI Knowledge Assistant
TechCrunch: Backstage access: Spotify’s dev tools side-hustle is growing legs
Spotify’s Backstage Blog: AI knowledge assistant and data plugins coming to Spotify Portal
Read what else we’re nerding out about on the Spotify Engineering Blog: ⁠engineering.atspotify.com⁠
You should follow us on Twitter ⁠@SpotifyEng⁠, ⁠LinkedIn⁠, and ⁠YouTube⁠!

Last year Spotify launched a big update to the app: a new typeface. 
For most of us, changing fonts is easy. It’s just a dropdown menu away. But creating a whole new typeface and then rolling it out across 45 unique platforms, and over 2,000 types of devices spread across 200 brands – that’s not so simple. 
This brand new font is called Spotify Mix and it was made just for Spotify. From playlists to microsites and billboards, it’s what you’ll see everywhere you see Spotify, representing the brand’s distinctive voice. 
In this episode, we’ll get into the technical and aesthetic challenges that go into creating and deploying a new typeface as well as what made its release possible: Spotify’s internal design system, known as Encore. 
Host and principal engineer Dave Zolotusky speaks with two people on Spotify’s design platform team who helped bring Spotify Mix to the world: an iOS engineer and “Spotify’s one and only typographer” — a designer who specializes in type and fonts.
Learn more about Spotify Mix and our internal design system, Encore:
Introducing Spotify Mix, Our New and Exclusive Font — Spotify
Creating coherence: How Spotify’s design system goes beyond platforms — Figma
Design Systems Podcast, Ep. 84: Digital typography: Suggesting, not dictating
Read what else we’re nerding out about on the Spotify Engineering Blog: engineering.atspotify.com
You should follow us on Twitter @SpotifyEng, LinkedIn, and YouTube!