Research Scientist - Generative AI Security

About Us

Pangea is a well-funded startup whose leaders and investors rank among the most successful veterans of the security industry. Pangea provides developers and security teams the industry’s Security Platform as a Service that uniquely empowers developers to build Secure by Design applications faster.

We live and breathe our company values:

Team - Every team member relies on each other, and we build a strong bond through open and transparent communication. By doing this well, we deliver for our customers, and they can confidently rely on us. Trustworthy teams are how we can build an incredible company together.

Culture - Our team is best when every individual brings their own uniqueness with confidence. We never stop learning and never stop questioning. Every interaction with our teammates or customers is an opportunity to learn something. Work hard and have a great time while we accomplish great things together.

Innovation - Speed matters. We obsess over our customers and react to their needs quickly to give them an exceptional experience from the products that we build. We do this by connecting our own builders directly with customers, taking calculated risks, and making good decisions quickly.

Leadership - We thrive on open communication and teamwork. Managers are at Pangea to manage complexity, and not control. Managers give their team the tools, the opportunities, and support needed to deliver, and provide them with the freedom to succeed. We win as a team.

We re seeking an exceptional Research Scientist to join our team at the forefront of Generative AI Security. This is an opportunity to shape the future of AI safety and security while working with cutting-edge large language models and multimodal systems in a fast-paced, innovative startup environment.

• Model Development:
  Train and fine-tune large language models (LLMs) and multimodal AI systems for security applications
• Research & Innovation:
  Conduct advanced research in AI security, developing novel approaches to protect against emerging threats
• Automation & Scripting:
  Build robust automation pipelines and tools to streamline model training, evaluation, and deployment processes
• Security Integration:
  Apply deep security domain knowledge to enhance AI model safety and reliability
• Cross-functional Collaboration:
  
  Work closely with engineering, product, and security teams to translate research into production solutions

• Advanced degree in Computer Science, Machine Learning, AI, or related field (PhD strongly preferred)
• Hands-on experience training and fine-tuning large language models and multimodal systems
• Strong programming skills with experience in automation and scripting (Python, PyTorch/Tensor Flow, etc.)
• Demonstrated expertise in Natural Language Processing and Machine Learning
• Experience with distributed computing and model scaling techniques

• Security Domain
  
  Experience:
  
  Background in cybersecurity, AI safety, or related security fields
• Publications in Speaking at top-tier ML/AI conferences
• Experience with model evaluation frameworks and safety benchmarks
• Knowledge of adversarial ML and robust AI systems
• Familiarity with MLOps and production ML systems

• Startup Mindset:
  Thrives in a lightning-fast paced environment with rapidly changing priorities
• Innovation Drive:
  Passionate about pushing the boundaries of what s possible in AI security
• Technical Excellence:
  Commitment to rigorous research methodology and high-quality implementations
• Collaborative Spirit:
  Enjoys working in cross-functional teams and mentoring others

• Cutting-Edge Impact:
  Lead groundbreaking research that directly influences the future of AI security
• Rapid Growth:
  Be part of a team that s defining an entirely new category in the AI space
• Technical Freedom:
  Work with state-of-the-art resources and have significant autonomy in your research direction
• Mission-Driven:
  Help build the secure AI infrastructure that the world needs

• Access to substantial compute resources for large-scale model training
• Collaborative research environment with publication…