Terminal Bench Expert

Company Description

MillionLogics, a trusted Oracle Partner, is a global IT solutions leader with a presence in London, UK, and a development hub in Hyderabad, India. Specializing in transformative technologies, the company empowers organizations through Data & AI services, Cloud migrations, and enterprise application optimization, with a strong focus on Oracle Cloud and database technologies. With a dedicated team of over 55+ AI experts, MillionLogics tailors cutting-edge IT solutions to drive tangible outcomes for clients. Guided by a commitment to innovation and excellence, MillionLogics delivers strategic IT consulting, custom application development, and security architecture solutions, among other offerings, to help businesses unlock their full potential. Discover more about their team and services at: millionlogics.com.

Role Description

This is a contract-based remote position for a Terminal Bench Expert. We are looking for highly analytical engineers, researchers, and domain specialists to contribute benchmark tasks for AI agent evaluation systems (e.g., Terminal-Bench). Design realistic, technically deep tasks simulating real-world scenarios such as debugging, data corruption, infrastructure failures, and complex workflows.

Offer Details:

• Mode of work: Fully Remote
• Pay: INR 1.25 to INR 2 lakhs per month (net/take-home)
• Duration: 12 months (likely extended)
• Experience: 3-10 years
• Number of positions: 28
• Evaluations: 1 round of technical interview

What does day-to-day look like:

• Design high-quality Terminal-Bench task ideas and specifications.
• Develop complex tasks requiring reasoning, investigation, and debugging.
• Write clear task descriptions, solution approaches, and verification logic.
• Define deterministic, outcome-based evaluation criteria.
• Identify realistic failure modes, edge cases, and operational constraints.
• Create tasks that challenge AI systems while remaining solvable by experts.
• Collaborate with reviewers to refine task quality and difficulty.
• Contribute expertise across one or more specialized domains.

Required Skills:

• 3–10 years of experience in software engineering or relevant domains.
• Strong debugging, reasoning, and analytical skills.
• Good understanding of system design, workflows, and dependencies.

  Looking to get Placed? Try our Placement Guarantee Plan

  Get Hired
• Ability to analyze complex systems across multiple layers.
• Experience with production systems, pipelines, or large-scale workflows.
• Strong technical writing and documentation skills.
• Exposure to LLMs, agentic systems, or AI evaluation frameworks.
• Experience reviewing technical specifications or designing validation logic.

Additional Details:

• Commitments Required: 40 hours per week with overlap of 4 hours with PST
• Employment type : Contractor assignment (no medical/paid leave)

How to apply

Please send us your updated CV to [HIDDEN TEXT] with email subject: TERMINAL BENCH