SwarmBench Task Engineer (Remote)

About the role

Turing is seeking a SwarmBench Task Engineer to design and build complex multi‑agent benchmark tasks that emulate real‑world planning, scheduling and operational decision‑making. This remote contractor role focuses on creating constraint‑rich problems to evaluate multi‑agent reasoning, decomposition and optimization capabilities.

Key responsibilities

• Design and develop multi‑agent benchmark tasks covering planning, scheduling and resource allocation.
• Create constraint‑rich problem statements with interacting variables and dependencies.
• Develop verification scripts to assess feasibility, completeness and optimality of solutions.
• Build decomposition strategies that split tasks across specialized sub‑agents (resource‑based, conflict resolution, optimization).
• Model realistic operational scenarios with timelines, resource constraints and incident response.
• Collaborate with the team to improve task quality, coverage and evaluation rigor.

Required profile

• 5+ years of experience in operations, project management, logistics or supply chain.
• Strong ability to formalize constraints, dependencies and scheduling logic.
• Proficiency in Python for building verification and validation scripts.
• Clear and precise technical writing skills.
• Experience with AI coding benchmarks such as SWE‑bench or Terminal‑bench.
• Hands‑on experience with Docker (Dockerfiles, image builds, debugging).

Required skills

• Python
• Docker

What we offer

• Fully remote work environment.
• Opportunity to contribute to cutting‑edge AI projects with leading LLM companies.
• 40‑hour work week with a 4‑hour overlap in PST.
• Freelance contract (initially 4 weeks, adjustable based on engagement).