Sauna.Ai
Senior Applied AI Engineer
San Francisco. Posted 3 months ago
The role
You will build, refine, and scale agent systems inside Sauna from architecture to deployment. You will work across infra, frontend, and product to ensure agents work in production.
What you will do
- Build, refine, and scale agent systems from architecture to evals to deployment
- Implement multi-step, tool-using agents that hit real APIs
- Design agent memory systems that persist relevant state
- Create agents that proactively do work and send reminders
- Own and evolve the eval framework
- Fix production failures from traces
- Pick agent problems worth solving without waiting for a ticket
- Make architecture decisions with incomplete information
- Decide what counts as reliable for agents
- Review other people's code and design docs
Required qualifications
- 4+ years of engineering experience, including time shipping production software
- Built and deployed agent-like systems: multi-step LLM pipelines, tool-using bots, scripted assistants, or similar
- Ran at least one agent system in production and made the architecture calls on it
- Hands-on experience with agent orchestration and memory management
- Hands-on experience with tool use and orchestration
- Hands-on experience with evaluation: success metrics, regression testing, and improving agent behavior
- Write production-grade code and work across systems without needing a spec
Nice to have
- Shipped agents that live in the wild, used by customers
- Familiarity with LLM ops, tracing, observability, and failure handling
- Been a founder or early engineer
- Mentored or led other engineers
Benefits
- health
- dental
- 401(k)
- considerable PTO
- gym budget
- lunch
Before you apply
- Must be based in the Bay Area or willing to relocate before starting
- Require US work authorisation, but open to O-1 or J-1 visa sponsorship for exceptional candidates
- Work trial is paid, in-person, typically 5 days
About the company
Sauna is a hybrid of AI and software. Wordware is the team behind it, 15 people backed by Spark Capital, Felicis, and Y Combinator.