Lead Software Engineer (Safety)

Overview

In this role you will shape the technical strategy and architecture for secure, production-grade AI systems that operate in high-stakes domains. You’ll bridge AI research and production, setting engineering standards and driving scalable ML infrastructure. As a senior advisor, you’ll mentor teams and guide cross-functional partners on safe, ethical AI deployment. You’ll collaborate with Frontier Labs to advance safety evaluations and governance. This is a high-impact opportunity to influence real-world AI applications at scale.

Responsibilities

• Steer technical strategy, architecture, and rigorous capability testing for complex ML systems
• Architect production-grade, secure ML software, orchestration tools, and scalable cloud infrastructure
• Establish company-wide standards for safe, reliable, and ethical ML deployment at scale
• Collaborate with cross-functional teams and senior stakeholders to translate client needs into resilient AI solutions
• Advise senior clients and partners as a trusted AI safety and technical authority
• Mentor and develop engineering talent, fostering an inclusive, high-performing culture
• Partner with Frontier Labs to advance safety evaluation procedures, multi-agent tooling, and governance frameworks

Key requirements

• Leadership experience designing, deploying, and securing scalable LLM applications, multi-agent frameworks, and AI safety evaluation processes
• Exceptional software engineering fundamentals with advanced Python expertise
• Hands-on experience with cloud platforms (AWS, Azure, GCP), scalable architectures, Docker, Kubernetes, agent sandboxing, and end-to-end cybersecurity
• Mastery of the full ML lifecycle with production systems using PyTorch, TensorFlow, or similar frameworks
• Autonomy, strategic ownership, and a track record of guiding technical decisions
• Strong communication with technical teams and senior non-technical stakeholders

• leadership
• strategic thinking
• effective communication
• Python (advanced)
• LLM architectures
• multi-agent frameworks

…