TL;DR — You own how we measure and improve Lovable's AI agent. You build the eval systems and experiments that tell us whether a change makes the agent better or worse, and you turn agent telemetry into the fixes that raise success rates and cut errors.

At Lovable, data scientists are not isolated model-builders; they sit close to the product, experimenting continuously with how intelligence changes user behavior and product dynamics.

Why Lovable?

Lovable lets anyone and everyone build software with any language. From solopreneurs to Fortune 100 teams, millions of people use Lovable to transform raw ideas into real products, fast. We are at the forefront of a foundational shift in software creation, which means you have an unprecedented opportunity to change the way the digital world works. Lovable-built applications and websites are visited hundreds of millions of times a month, and our enterprise footprint is compounding fast. And we're just getting started.

We're a small, talent-dense team building a generation-defining company from Stockholm. We value extreme ownership, high velocity, and low-ego collaboration. We seek out people who care deeply, ship fast, and are eager to make a dent in the world.

What We're Looking For

A data scientist who wants to make an AI agent measurably better, not just report on it. You own agent quality metrics and drive them up.
Experience or strong interest in LLM evaluation and observability: building evals, scoring outputs, tracing agent behavior, and catching regressions.
Strong SQL and Python, applied statistics, and experimentation. Comfortable designing A/B tests for agent changes where outcomes are noisy.
You build systems and agents that produce this insight continuously, rather than one-off analyses.
Instinct for what "good" looks like in agent behavior (success, error rates, task completion) and how to measure it when there is no clean answer key.
Entrepreneurial. Thrives in ambiguity, and works closely with the engineers building the agent.

What You'll Do

Define and own the metrics for agent quality: success, completion, error rates, and the behaviors that drive them.
Build the eval systems and experiment framework that decide whether an agent change ships, like an A/B-tested rollout that catches a change increasing errors before it reaches everyone.
Turn agent traces and telemetry into concrete fixes, working directly with the agent engineering team.
Build the tooling and agents that produce these evaluations continuously as the agent evolves.
Set the bar for how we judge agent behavior where there is no answer key to check against.

Our tech stack

We're building with tools that both humans and AI love: