Write clear and accurate problem descriptions in both Chinese and English, defining precise input/output formats and constraints.
Design comprehensive and robust test datasets, including edge cases and extreme scenarios, to ensure accurate evaluation and prevent cheating.
Write reference solutions and automatic validation logic, and perform difficulty grading and quality assessment.
Participate in team cross-reviews to ensure the scientific validity, fairness, and innovation of the benchmark.
Required skills
Experience participating in or creating problems for programming competitions like ACM/ICPC, Codeforces, etc.
Expertise in algorithms and data structures, with a specialization in at least one area such as graph theory, dynamic programming, or string algorithms.
Proficient in Python or C/C++ for writing efficient algorithm implementations and validators.
Excellent problem abstraction skills to transform vague ideas into unambiguous algorithmic problems.
Preferred skills
Understanding of the capability boundaries and common pitfalls of major Code LLMs.
Experience writing algorithm blogs or teaching courses.
Experience in benchmark design or contributions to open-source Online Judge (OJ) systems.
Skilled in designing complex problems that test the logical reasoning abilities of AI models.
Experience and education
Background in Computer Science, Software Engineering, Mathematics, or Artificial Intelligence; or experience as an Algorithm Engineer or Software R&D Engineer (Algorithm-focused).
Experience in evaluating code generation models is a plus.
Thriveth makes AI data-training work easier to find, understand, and navigate. We replace uncertainty with clear opportunities, realistic expectations, and insights from real application journeys.
Hiring companyRemoExperts
Turing