Software Engineer – Terminal Bench (Remote) | Immediate Start

Company Overview

Dusker AI is an innovation-driven company focused on advancing AI systems through high-quality data, evaluation, and benchmarking. We develop large-scale benchmarks such as Terminal-Bench and build automated workflows that rigorously test the real-world capabilities of AI agents.

Our mission is to enhance the reliability, accuracy, and robustness of AI models by creating challenging, reproducible, and research-grade evaluation environments. With strong expertise in systems engineering, automation, and developer tooling, we help organizations build AI agents that perform consistently in practical terminal-based environments.

We are expanding quickly and looking for individuals who can contribute immediately and grow with us.

Role Description

We are seeking a highly skilled Software Engineer (Terminal-Bench) for a 3-month contract, with a strong likelihood of extension based on performance and project needs.

This position is fully remote and starts immediately.

In this role, you will:

• We are seeking a highly skilled Software Engineer (Terminal-Bench) for a 3-month contract with a strong possibility of extension.
• This is a fully remote role where you will work on designing, maintaining, and improving Terminal-Bench tasks, automation pipelines, and evaluation workflows.
• You will work extensively in Linux/Ubuntu environments, write efficient Bash scripts, debug complex terminal behaviors, and manage Docker-based execution setups.
• The role requires independently analyzing issues, improving task reliability, documenting processes, and collaborating with remote engineering teams to enhance the performance and accuracy of benchmarking systems.

Qualifications

• Strong proficiency in Linux/Ubuntu command-line operations
• Expert-level Bash scripting and terminal automation
• Hands-on experience with Docker containers, environment setup, and isolation workflows
• Understanding of system calls, processes, filesystems, and core OS behaviors
• Excellent debugging and problem-solving abilities
• Ability to work efficiently in a fully remote environment with strong attention to detail

Preferred Skills

• Familiarity with Terminal-Bench, OSWorld-SFT, or similar evaluation ecosystems

Compensation

India: ₹12–15 LPA

International: $8–$10 per hour

Employment Type

Contract — 3 Months

High potential for extension

Location

• Remote — Global