Collaborate With Us

Join us in advancing the understanding of long-term AI agent behavior and making AI systems safer through rigorous evaluation.

Who We're Looking For

We're keen to collaborate with various stakeholders in the AI ecosystem who share our commitment to long-term AI safety.

AI Labs & Safety Teams

Organizations developing AI agents who want to run extended evaluations on their systems and understand long-term behavior patterns.

  • • Multi-day evaluation testing
  • • Goal drift detection
  • • Safety assessment frameworks
  • • Collaborative research studies

Researchers

Individual researchers interested in designing or contributing to new evaluation scenarios and metrics for long-term agent behavior.

  • • Novel evaluation scenarios
  • • Metric development
  • • Collaborative publications
  • • Open-source contributions

Policy & Governance

Policy entities looking for empirical data to support or craft guidelines on safe deployment of autonomous AI systems.

  • • Empirical safety data
  • • Risk assessment insights
  • • Regulatory framework input
  • • Policy recommendations

Collaboration Opportunities

Research Partnerships

Collaborate on empirical studies, develop new evaluation frameworks, or contribute to our understanding of long-term agent failure modes.

  • • Joint research projects
  • • Co-authored publications
  • • Shared datasets and findings
  • • Workshop and conference presentations

Technical Collaboration

Help improve our open-source tools, contribute evaluation scenarios, or integrate our frameworks with your existing systems.

  • • Framework development
  • • Tool integration
  • • Evaluation scenario design
  • • Code contributions

Evaluation Services

We can help organizations run extended evaluations on their agent systems using our frameworks and expertise.

  • • Custom evaluation design
  • • Multi-day testing protocols
  • • Behavior analysis and reporting
  • • Safety assessment consultation

Knowledge Sharing

Participate in workshops, contribute to our documentation, or help establish best practices for long-term agent evaluation.

  • • Best practices development
  • • Community workshops
  • • Educational content creation
  • • Standard setting initiatives

How to Get Involved

1. Reach Out

Contact us via email to discuss your interests and how we might collaborate. We're happy to explore custom arrangements that work for your organization.

2. Explore Our Tools

Check out our Long Agent Framework on GitHub and see how it might fit with your existing evaluation processes or research goals.

3. Join the Conversation

Participate in discussions about long-term AI safety, contribute to our research insights, or help shape the future of agent evaluation standards.

Get in Touch

Ready to collaborate? We'd love to hear from you and explore how we can work together to advance long-term AI safety.

Email

Reach out directly to discuss collaboration opportunities, research partnerships, or any questions about our work.

diogo.abc.cruz@gmail.com

GitHub

Contribute to our open-source framework, report issues, or explore our codebase and evaluation tools.

View Framework

Response Time

We typically respond to collaboration inquiries within 2-3 business days. For urgent matters, please indicate the priority in your subject line.

Ready to Make AI Safer?

Join us in building the tools and knowledge needed to ensure AI agents remain safe and aligned over extended periods of operation.