AI Engineer
Zyphra · San Francisco
CHECKED 6H AGO
Frontier labonsite
At a glance
Highlights
- Build a next-gen desktop and browser agent that autonomously completes tasks.
- Work across browsers, operating systems and enterprise filesystems at scale.
- Push the boundary of what production-grade software agents can do.
What they offer
- Comprehensive medical, dental, vision, and FSA plans
- Competitive compensation and 401(k) plan
- Relocation and immigration support on a case-by-case basis
- In-office snacks and meals provided
- Unlimited PTO and company holidays
Ideal candidate
A fast-moving, curious builder driven to ship production-grade ML agents. Thrives in ambiguity, values bold ideas, and collaborates closely across product, research, and engineering.
Requirements
- Proficiency in Python and a deep understanding of building and debugging complex ML-driven applications
- Experience working with desktop operating systems (Windows and macOS), including APIs for screen reading, file interaction, and accessibility frameworks
- Experience developing browser extensions or automation tools with fine-grained control over the browser (mouse, tabs, DOM)
- Understanding of LLMs, prompting techniques, and orchestration frameworks for multi-step reasoning
- Ability to work across the full ML stack, from model integration to serving infrastructure
- Experience designing or working with secure and virtualized execution environments
- Excellent communication and collaboration skills across product, research, and engineering teams
- remote policy
- In-person team in San Francisco
- visa sponsorship
- Immigration support on a case-by-case basis
Lab profile
About Zyphra
San Francisco AI lab building open language and audio models and autonomous desktop agents.
What they build
- Desktop and browser-based autonomous agents
- Open language models
- Open-source audio models (ASR, TTS, speech-to-speech)
Known for
- Agentic systems and human-AI interaction
- Reinforcement learning, post-training and preference learning
- Large-scale language model pre-training
- Novel model architectures and training methods
Key facts
- Headquartered in San Francisco, California