Founding Backend Engineer

Mozart AI; Greater London; On-site; GBP 90,000 - 130,000; Full time

Job summary

Mozart AI in London is seeking a backend-focused engineer to own APIs, design the media generation backbone, and run media at scale across audio, video, and images. You will shape the core data model for DAW sessions, social graphs, and collaboration features, while supporting payments and observability.

This in-person role offers the chance to impact billions of potential music creators and fans as the platform evolves weekly, with high ownership and fast delivery expectations.

Qualifications

• Experience designing contracts used by web, mobile, and internal tools.
• Built multi-service workflows with clear failure modes.
• Modelled complex, evolving data domains with history and concurrency.
• Experience with media processing pipelines and real-time collaboration.

Responsibilities

• Own the API surface and end-to-end contracts for multiple clients.
• Build the media generation backbone across providers and integrations.
• Manage media at scale: upload, transcoding, storage, and delivery.
• Design relational, versioned data models across sessions and social graphs.
• Ship real-time collaboration features with presence and shared state.
• Own payments, subscriptions, credits, and refunds.
• Make the platform observable with traces, logs, and metrics.
• Develop CI/CD for agents and ensure reliable deployments.
• Design for constant change with zero-downtime migrations.

Skills

APIs design

Distributed systems

Domain modeling

Media pipelines

Billing & subscriptions

AI-native tooling

Ownership & velocity

5+ years experience

Tools

CRDTs

Job description

Mozart AI is music's first creation network, where the world goes from listening to music, to making it together, to being known for it. 1.5 billion people can play an instrument. We're building the products that turn them into artists, and turn their fans into the next ones. Our customers love us, and include everyone from grandmas using Vibe sessions to Grammy winners using the Studio. Music is having its first explosive moment of creation-driven consumption, and we aspire to shape that world to preserve the social and fun elements of music creation.

We've raised over $7M from Balderton Capital (backers of Revolut) and several Silicon Valley legends like Kevin Hartz and Emery Wells, who have backed companies like Uber, Spotify, Reddit, and other massive consumer outcomes.

We're a small, talent-dense team building the future of music from our London office. We operate with high ownership, high velocity, and low-ego collaboration — specialist generalists who work hard, learn fast, and ship relentlessly. We care deeply, move fast, and we're here to build something that outlives us.

About the Role

• Mozart combines music creation, real-time collaboration, and a creator community in one product, and the backend has to hold all of it together as usage scales. You'll build the backend foundation that carries Mozart from hundreds of thousands of users to millions, and you'll do it while the product changes every week.
• This is an intense, in-person role, built for people who want massive impact and grow their careers faster than anywhere else.

What You'll Do

• Own the API surface. Design the typed, end-to-end contracts that web, mobile, and internal admin tooling consume
• Build the media generation backbone. Design how every AI job moves from request to result across many internal and external providers and integration styles, and how the steps that follow chain together without ever losing a job. The result reaches every collaborator in the session, not just the one executing the job.
• Run media at scale. Upload, transcoding, storage, and CDN delivery for audio, video, and images
• Design the core data model. From the state of a DAW session (projects, songs, tracks, clips) to the social graph around it (profiles, follows, feeds, publishing, moderation), build schemas that are deeply relational, versioned, and consistent across every client.
• Ship real-time collaboration. Presence, shared state, and conflict resolution for multiple people editing the same session at once
• Own payments and monetization. Stripe and RevenueCat subscriptions, pay-as-you-go credits, gifting, refunds, and the edge cases between platforms.
• Make the platform observable and safe. Traces, logs, and metrics for running the system, plus the durable data we use to analyse and evaluate the quality of our generations. Guardrails on top for cost and abuse too.
• Build CI/CD for agents. Per-PR environments, seeded fixtures from a single user to a full collaborative session, and test infrastructure that lets automated suites and AI agents verify changes before merge.
• Design for constant change. Set up how we allow everyone in the team to constantly iterate on the user experience at high speed: zero-downtime migrations, coordinated client and server releases, constantly evolving APIs and schemas, cleaning up discipline.

What We’re Looking For

• APIs with multiple consumers. You have designed and maintained contracts used by web, mobile, and internal tools on different release cycles, and you know the quirks each platform brings.
• Multi-step, multi-service systems. You have built workflows that span queues, callbacks, internal and external providers, and you have a clear model of where such systems break.
• Complex domain models on live data. You have modelled a deeply relational domain with history and concurrent edits, changed its schemas, storage, and API shapes while hundreds of thousands of users were on it, and removed the old path afterwards.
• Media pipelines. You have built upload, processing, and delivery for audio or video at a scale where storage and bandwidth were real costs.
• Billing and entitlements. You have worked on subscriptions, credits, or refunds and understand the failure modes around charging and crediting.
• AI-native. You work with coding agents daily and design environments and tests so an agent can verify a change on its own.
• Ownership, velocity, low ego. You care about outcomes after release, not just merges. You challenge what doesn't make sense, but also update your thinking when new evidence comes in, and do whatever the team needs.
• 5+ years of professional experience

Nice-to-Haves

• Real-time collaboration systems (CRDTs, presence, shared state)
• Per-PR environments or episod... It should be "Per-PR environments or episod..."
• Building evaluation datasets or quality pipelines for generative media
• Social feeds, search, or recommendation systems
• Server-side rendering and multi-platform delivery