Reinforcement Learning Engineer

About the role

This position is responsible for creating the foundational systems that generate the training environments required for advanced AI models used in cybersecurity research. You will own the design and implementation of the infrastructure that translates complex real-world security problems into large-scale learning scenarios for next-generation AI agents. Your contributions will directly shape how these AI systems learn to identify, analyze, and defend against software vulnerabilities in a simulated setting. The role sits at the intersection of distributed systems, security research, and machine learning to solve novel engineering challenges. You will collaborate closely with researchers to ensure the environments you build are rigorous, measurable, and aligned with scientific goals. This position requires a strong bias for building robust and scalable systems that can handle demanding computational workloads over time. You will be a key architect in defining the standards for how AI training data is generated and processed for security-specific tasks.

Key facts

What you'll do

Design and build data processing pipelines that ingest software projects and convert them into structured reinforcement learning environments.

Develop and maintain the core infrastructure that powers AI-driven cybersecurity research and experimentation.

Operate at the intersection of artificial intelligence, security analysis, and systems engineering to solve complex technical constraints.

Construct scalable simulation environments that enable AI models to learn critical skills such as vulnerability discovery and safe exploitation.

Implement low-level debugging procedures to diagnose issues within complex containerized and virtualized testbeds.

Work with large open-source codebases, navigating complex build systems to instrument and modify software for learning purposes.

Create tooling that automates the setup, configuration, and teardown of reproducible Linux environments for training episodes.

Partner with research teams to translate theoretical security problems into executable benchmarks that an AI agent can interact with.

Ensure that the generated environments are secure, isolated, and capable of producing consistent and verifiable learning signals.

Optimize the performance of environment execution to support high-throughput training runs across distributed compute resources.

Document system behaviors and interfaces to facilitate collaboration with engineers, scientists, and security experts.

Contribute to the evolution of the platform by integrating new tooling, languages, and analysis techniques as the research scope expands.

Requirements

Demonstrate a clear understanding of reinforcement learning training processes, including environment interaction, state observation, and reward shaping as used in modern AI systems.

Possess hands-on experience building and managing reproducible Linux environments using container technologies such as Docker and containerd.

Showcase strong proficiency in both Python and C programming languages, with the ability to write efficient and maintainable code in each.

Exhibit familiarity with common software vulnerabilities, fuzzing techniques, or program analysis methods relevant to security research.

Have practical experience with build systems and the ability to work effectively within large, complex open-source codebases.

Be comfortable operating within Linux terminal environments and performing low-level debugging using system tools and logs.

Maintain a strong commitment to best practices in software engineering, including version control, testing, and code review.

Be prepared to work in a fast-paced environment where priorities can shift based on research needs and emerging security challenges.

Nice to have

Bring additional value to the team with experience in DevOps pipelines, such as GitHub Actions, to automate testing and deployment workflows.

Show familiarity with advanced build tools like BuildKit or Nix to manage complex and reproducible build processes.

Demonstrate proficiency in the Rust programming language, particularly for systems-level programming and performance-critical components.

Have direct experience working with benchmark environments such as Capture The Flag (CTF) challenges or the SWE-bench framework.

Practical notes

Salary range: $176,400 - $242,550 annually.

This role may be eligible for a discretionary bonus or commission plan based on performance.

Bugcrowd is an Equal Opportunity Employer and values diversity.

Accommodations are available for individuals with disabilities during the application and interview process.

Please submit your application through the careers page at bugcrowd.com/about/careers.