Applied Scientist, Frontier Cybersecurity AI

Mistral Ai; Paris; On-site; EUR 80,000 - 120,000; Full time

Job summary

Mistral AI is seeking Applied Scientists to build frontier cybersecurity models. You will develop SOTA models across cybersecurity domains, from vulnerability analysis to incident response and remediation.

You’ll create data pipelines, training environments, and evaluations that enable measurable security capabilities, working with cross-functional teams and production data. We are researchers, engineers, product builders, and former founders who love tackling complex real-world problems;

Qualifications

• Fluent in English with excellent communication.
• Experience building harnesses, benchmarks or RL environments for code, agentic or security tasks.
• Hands-on security background in vulnerability research, exploit development, red teaming, or detection and incident response.
• Not afraid of contributing to a big codebase; works independently with little guidance.
• Writes clean, readable, high-performance, fault-tolerant Python code.
• Doesn’t need roadmaps: moves independently and navigates uncertainty with confidence.
• Low-ego, collaborative and eager to learn.
• Track record of success through personal projects, professional projects or academia.

Responsibilities

• Develop, pre-train and post-train domain-specialized foundation models for agentic defensive security.
• Work across a diverse set of complex real world problems to build the future of AI for cybersecurity.
• Design evaluations and benchmarks that measure realistic security capability on customer-representative tasks.
• Build sandboxed execution environments, RL tasks, and verifiers grounded in real customer workflows, and generate and curate the training data for training frontier models.
• Develop the necessary tools and frameworks to facilitate data generation, model training, evaluation and deployment.
• Own technical delivery across project sprints and manage research communications with customer teams.

Skills

English communication

Security domain experience

Python

Independent problem solving

Codebase navigation

Team collaboration

Project track record

Education

PhD / Master in Mathematics, Security or ML

Tools

Python

Job description

Mistral AI is seeking Applied Scientists to build frontier cybersecurity models. You will develop SOTA models across cybersecurity domains, from vulnerability analysis to incident response and remediation.

You’ll create data pipelines, training environments, and evaluations that enable measurable security capabilities, working with cross-functional teams and production data. We are researchers, engineers, product builders, and former founders who love tackling complex real-world problems;