Software Engineer

Clorvin Labs; United States; Remote; USD 90,000 - 130,000; Full time

Job summary

Clorvin Labs in the United States is seeking a developer to design terminal-based tasks that mirror real-world software engineering challenges. You will build and maintain Docker environments, creating reference solutions and validation tests to assess AI models.

You will run tasks against state-of-the-art models, analyze results, and refine tasks to improve evaluation accuracy. Candidates should have a CS/engineering degree, strong Linux and Python skills, and a track record solving genuine

Qualifications

• Degree in Computer Science, Engineering, or related discipline (B.Tech, M.Tech, or PhD).
• Proficiency in Linux and hands-on experience with Docker.
• Fluency in Python or another programming language.
• Track record of tackling and solving genuine technical problems.
• Familiarity with Terminal-Bench or comparable AI evaluation frameworks is valued, though not required.

Responsibilities

• Design complex terminal-based tasks that mirror challenges software engineers face in practice.
• Build and maintain Docker environments that reliably reproduce the problems these tasks present.
• Create reference solutions and validation tests that catch when an AI model's answer falls short.
• Run tasks against state-of-the-art AI models, analyze their performance, and refine based on the results.

Skills

Linux

Docker

Python

Problem solving

Test automation

Education

B.Tech / M.Tech / PhD in CS/Engineering

Tools

Terminal-Bench

Job description

Batch: 2020/2021/2022/2023/2024.

About the role

• You will design complex terminal-based tasks that mirror challenges software engineers face in practice
• Build and maintain Docker environments that reliably reproduce the problems these tasks present
• Create reference solutions and validation tests that catch when an AI model's answer falls short
• Run your tasks against state-of-the-art AI models, analyze their performance, and refine based on the results

What you'll need

• A degree in Computer Science, Engineering, or a related discipline at B.Tech, M.Tech, or PhD level
• Proficiency in Linux and hands-on experience with Docker
• Fluency in Python or another programming language
• A track record of tackling and solving genuine technical problems
• Familiarity with Terminal-Bench or comparable AI agent evaluation frameworks is valued, though not required

What you'll bring

• The ability to break down real-world engineering work into reproducible scenarios
• Skill in scripting and testing to verify correctness
• Attention to detail in building environments and solutions that withstand scrutiny from advanced systems