Fine-Tuning AI Jobs

Lead, develop, and grow the managers and teams responsible for the consumer harms portfolio, including child safety, user well-being, harmful manipulation, and election integrity.

Design and run experiments to improve Claude's perception and agentic capabilities. Develop robust, reliable evaluation frameworks for measuring our models' ability to complete complex computer tasks.

Define behavioral defaults and steerability constraints. Develop and maintain taxonomies of model behaviors across capabilities. Identify, triage, and prioritize behavior issues and opportunities, coordinating input from.

Prioritize the team’s work in collaboration with the technical lead, research teams, and product teams to support fast iteration on research projects and training runs.

Build and operate data pipelines that extract data from research training runs and land it in storage systems that are easy and fast to query.

Build and extend web platforms for RL environment creation, management, and quality review — including environment configuration, versioning, and validation workflows.

Our team is a quickly growing group of committed researchers, engineers, policy experts, and business leaders working together to build beneficial AI systems.

Design, build, and run capability evaluations to assess what new models can do in the biological domain, and turn results into concrete deployment recommendations.

Implement and optimize post-training techniques at scale on frontier models. Conduct research to develop and optimize post-training recipes that directly improve production model quality.

Lead, coach, and grow a high-performing team of AI and software engineers, developing them into strong technical leaders while fostering a culture of ownership, technical excellence, experimentation, and continuous learning.

Gaining a deep understanding of both our technology and our customers. Frontliners operate across a broad spectrum of responsibilities, much like a startup CTO. They work in small teams to own the end-to-end execution of high-stakes projects.