Machine Learning Research Engineer

Machine Learning Research Engineer

About Us
We’re Unitary, a fast-growing AI startup building Virtual Agents that combine deterministic code, LLM reasoning, and human expertise to automate complex operational workflows. Our technology powers trust & safety systems at scale, analysing millions of images and videos daily for enterprise customers.

Backed by Creandum and Plural, with over $25M raised, we’re scaling rapidly and shaping a future where AI can make safe, fair, and reliable decisions online.

The Role
We’re looking for a Machine Learning Research Engineer to pioneer how Virtual Agents are built, deployed, and scaled. You’ll design an “Agent Factory” — a system that can automatically generate agents from captured workflows, combining the speed of code with the reasoning power of AI.

Working closely with platform and ML teams, you’ll research, prototype, and productionise capabilities that let AI autonomously build reliable software workflows.

What You’ll Do
Design and build the Agent Factory to generate and deploy Virtual Agents
Develop code-generation and evaluation frameworks for workflow automation
Integrate LLM-based reasoning with deterministic Python systems
Benchmark and iterate on automation quality through experimentation
Collaborate with engineers and customer teams to deploy at scale
Contribute to Unitary’s ML research community and technical direction

About You
Expert in Python and experienced with LLMs, Agentic AI, or code-generation systems
Strong engineering fundamentals (software design, testing, DevOps, MLOps)
Curious and inventive — comfortable moving between research and implementation
Collaborative communicator who thrives in a small, fast-moving team
Bonus: experience with workflow orchestration (e.g., Temporal), browser automation (Playwright), CI/CD, Terraform, or scaling ML systems in production.

Why Unitary
Remote-first across Europe/North America
Competitive salary + equity
Flexible hours and generous parental/sick leave
Annual budgets for learning and wellbeing
3 annual team off-sites in London or Europe

Join us to push the frontier of agentic automation — where AI learns to build and operate systems that make the internet safer.