We are building the harness that makes powerful AI models usable for real operational work. The challenge is not the AI itself. The challenge is making it reliable enough that companies can trust it to run autonomously, recurrently, at scale.
You will constantly zoom out and zoom in. Understand the whole system, then dive into specifics. Balance capability with reliability. Balance cost with performance. Balance the generalist nature of probabilistic systems with the guardrails needed for production use.
What You Will Do:
- Build orchestration systems that decompose complex tasks across multiple agents
- Design the guardrails and configuration layer that makes powerful models enterprise-ready
- Make architectural decisions that balance cost, performance, reliability, and capability
- Ship production systems, not research prototypes
Requirements:
- You are obsessed with AI. You use LLMs daily. You see where autonomous agents are heading and want to build that future. If you do not work with AI every day, do not apply.
- You believe operational work should be automated so people can focus on meaningful activities
- Entrepreneurial mindset with a bias for action
- Deeply curious. You go deep, push hard, and learn continuously. Learning is your identity.
- Generalist who can zoom out to architecture and zoom in to implementation
- First-principles thinker who understands systems holistically
- Comfortable with ambiguity and probabilistic systems
- Can balance competing concerns: reliability vs capability, cost vs performance
Nice to Have:
- Built agent-based or agentic AI systems
- Experience making AI systems production-ready
What We Offer:
- Competitive salary and equity
- Remote-first, flexible hours
- Work at a high level of abstraction, AI handles the specialist work
- High autonomy, direct impact