The brief
Federato builds underwriting software for insurers, and it was hiring several forward-deployed machine learning engineers. The requirements pulled in two directions.
Candidates needed real production ML: deploying models and LLM workflows, and building the evaluation and monitoring around them, ideally at a startup rather than a big tech company. They also had to explain complex ML clearly to people who are not technical. The test the hiring team kept coming back to was whether they could picture the person in front of a customer.
What happened
Strong ML engineers who are also good with customers are rare, and job titles do not separate them from everyone else.
We submitted two candidates. One was hired and started in August 2026.