About the position
AI agents represent the next frontier, intelligent systems capable of independent action and complex decision-making. Their deployment promises unprecedented efficiency and innovation. But unleashing this autonomy creates an urgent need: how do we manage these agents effectively, ensure safety, and guarantee alignment with human values when agents operate in the wild? Misaligned or unmonitored agents aren't just a technical glitch; they're a fundamental barrier to trust and adoption. Scale AI's Oversight team is tackling this head-on, building the critical infrastructure for real-time monitoring, intervention, and control. We're defining how humans and autonomous AI coexist safely. Join us if you want to solve one of the most crucial problems in the path towards beneficial AGI.
Responsibilities
- Own large areas within our product
- Design and build the generalizable oversight platform applicable to various enterprise agents oversights applications
- Build features end-to-end: front-end, back-end, system design, debugging and testing
- Deliver experiments at a high velocity and level of quality to engage our customers
- Work across the entire product lifecycle from conceptualization through production
- Influence the culture, values, and processes of a growing engineering team
- Inspire and mentor less experienced engineers
- Collaborate with cross-functional teams to define, design, and ship new product features and experiences.
Requirements
- 5+ years of relevant experience is preferred
- Track record of shipping high-quality products and features at scale
- Desire to work in a very fast-paced environment
- Ability to turn business and product ideas into engineering solutions
- Excellent problem-solving skills, and be able to work independently or as part of a team.
- Excited to join a dynamic, hybrid team in either San Francisco or New York City
Benefits
- Comprehensive health, dental and vision coverage
- Retirement benefits
- Learning and development stipend
- Generous PTO
- Commuter stipend (may be eligible)
