About the team
Research keeps Vercilio on the frontier: evaluating new models the day they land, tuning our routing and reasoning settings, and turning messy real-world usage into measurable quality. We work close to the product, so research here ships.
About the role
You'll bridge research and engineering — running evaluations, analyzing real-world usage, and shipping improvements to how Vercilio selects and uses models.
Your work shows up in the product quickly: better routing, better defaults, better answers.
In this role, you will
- Build evaluation harnesses for new and existing models.
- Analyze usage data to find where quality breaks down.
- Prototype and ship improvements to routing and reasoning settings.
- Keep us current as the model landscape shifts week to week.
You might thrive here if
- Strong analytical skills and comfort with data.
- Solid software engineering — you can ship your own experiments.
- Familiarity with modern LLMs and their evaluation.
- A bias toward measurable, reproducible results.
Sound like you?
There's no formal application — just send us an email telling us a little about yourself, why this role, and anything you've built that you're proud of. We read every message.
Send us an email to join uscontact@vercilio.com