About the role

Senior engineer extending Menlo Security’s ingress proxy for AI-agent traffic. Building secure, high-performance systems for Menlo Security’s browser security platform.

Responsibilities

Join the team owning Menlo Security’s ingress proxy service, the primary path into the product for most users
Extend ingress proxy controls to cover AI-agent traffic
Own features from design through production
Optimize security and performance decisions affecting the system
Build and maintain production systems and asynchronous services
Safely uplift legacy systems while keeping them in production and moving traffic in stages
Refactor systems into more robust architectures and evaluate language trade-offs
Apply comprehensive testing and verification strategies
Collaborate with team members and cross-functional teams to understand customer needs, refine requirements, implement, review, and deliver changes
Provide technical leadership by articulating architectural decisions and sharing secure development practices

5+ years of experience building and operating production systems
Working knowledge of HTTP and TLS
Experience in Rust or Go strongly preferred
Solid hands-on experience building and maintaining production systems in Rust or Go
Demonstrated ability to optimize latency and throughput in networked applications using profiling tools and efficient concurrency models
Strong background implementing secure systems, including threat modeling, vulnerability mitigation, and data integrity in distributed environments
Experience building and operating asynchronous services handling high concurrency in production
Understanding of backpressure, cancellation, and non-blocking I/O failure modes
Experience uplifting legacy systems while they remain in production, moving traffic safely in stages, and refactoring into robust architectures
Experience with fuzzing, property-based testing, and integration tests
Ability to articulate and defend architectural decisions and share secure software development practices
Proficiency using LLMs thoughtfully and selectively while assessing outputs for correctness, maintainability, architecture fit, and engineering goals