Cryptography Specialist for AI lab

Responsibilities: — Design expert-level Terminal-Bench cryptography tasks by transforming real, paid work from your own experience into challenges solvable end to end by AI agents. — Define the task environment including files, systems, ciphertexts, and data to ensure the challenge is realistic, self-contained, and reproducible in a Linux command-line setup. — Develop working reference solutions in Python or SageMath that demonstrate task solvability, such as recovering keys or decrypting secrets. — Create automated checks and tests to objectively verify the correctness of final results and explain detection of mistakes or partial answers. — Calibrate task difficulty so that a typical expert requires about 60-75 minutes to solve it, maintaining challenge difficulty for AI agents while ensuring fairness and clarity. — Package each task with Docker and Git where applicable, provide clear documentation, and revise tasks based on reviewer feedback until quality standards are met. Requirements: — At least 3 years of hands-on cryptanalysis or applied cryptography experience; roles limited to GRC, audit-only, policy, sales, management, or coursework do not qualify. — Recent work within the last 3 years with ability to specify systems, tools, data, and rules used. — Proficiency with low-level tools such as SageMath and Python, and comfortable working on the Linux command line. — Ability to demonstrate findings with proofs such as decrypted secrets or recovered keys and explain error detection. — Technical fluency with files, CSV/JSON, APIs, or databases. — Availability to start around November 2026. Nice-to-have: — Published research or write-ups on cryptanalysis. — Strong Capture The Flag (CTF) results, especially in cryptography categories. — Discovery of CVEs in cryptographic implementations. — Experience with Docker, Git, and writing automated tests. — Proficiency in C for low-level cryptographic tasks. Benefits: — Work is task-based compensated, with payment per task. — Each task is expected to take approximately 60-75 minutes. — The application process includes completing a vetting task instead of a live interview. — Expected start date around early November 2026.