Open Source Contributor

Hyper Lychee Labs; India; Remote; INR 7,790,000 - 15,581,000; Part time

Job summary

Hyper Lychee Labs is seeking exceptional open-source contributors for an AI-training project involving complex technical workflows. You will create, solve, review, and validate tasks across software engineering, ML, science, hardware, or engineering domains.

Prior AI-training experience is not required; applicants with strong, publicly verifiable technical work are encouraged to apply. This is a contract, remote role with flexible hours.

Qualifications

• Strong programming ability with Python
• Meaningful contributions to open-source software
• Experience in software engineering or ML/data science is a plus
• Public evidence of skill through merged contributions or maintainer responsibilities
• Contributions to complex projects over extended periods are valued

Responsibilities

• Create, solve, review, and validate tasks based on real workflows
• Build reproducible task environments, reference solutions, tests, and evaluation criteria
• Review AI-generated work for correctness and quality
• Diagnose model failures and explain incorrect approaches or results

Skills

Python

Open-source contributions

Public repositories

Tools

Git

Job description

Pay: $60–$120/hour equivalent, paid per accepted task

Location: Global, fully remote

Job Type: Contractor

Time Commitment: 15+ hours per week

Schedule: Flexible

Job Description

We are looking for exceptional open-source contributors to support an AI-training project focused on complex, real-world technical workflows.

In this role, you will use your programming and domain expertise to create, solve, review, and validate challenging tasks in software engineering, machine learning, science, hardware, or engineering. Your work will help improve how advanced AI systems reason, write code, use technical tools, and complete long-horizon projects.

Prior AI-training experience is not required. We are primarily looking for candidates with strong, publicly verifiable technical work.

What You'll Work On

• Create realistic technical tasks based on professional, scientific, or engineering workflows.
• Build reproducible task environments, reference solutions, tests, and objective evaluation criteria.
• Review AI-generated work for correctness, completeness, and technical quality.
• Diagnose model failures and explain why an approach or result is incorrect.

Required Qualifications

• Strong programming ability, including practical proficiency with Python.
• Meaningful contributions to open-source software or substantial original public projects.
• Experience in at least one relevant area, including:
• Software engineering
• Machine learning or data science
• Scientific computing
• Hardware, embedded systems, or electronics
• Mechanical, aerospace, civil, chemical, or electrical engineering
• Developer tools, infrastructure, automation, or security

Public Proof of Skill

Applicants should be able to provide publicly verifiable evidence such as:

• Meaningful merged contributions to established repositories
• Maintainer or core-contributor responsibilities
• Substantial original repositories or technical tools
• Scientific or engineering software contributions
• Public code supporting published research
• Contributions involving implementation, debugging, testing, performance, infrastructure, simulation, or data processing

Forks, tutorial-only repositories, superficial documentation edits, contribution counts, or AI-generated repositories alone are not sufficient.

Preferred Qualifications

• Experience designing technical tests, benchmarks, evaluation tasks, or reproducible environments.
• Experience reviewing code or technical contributions from others.
• Competition results, publications, public technical writing, or other evidence of expertise.
• Previous experience contributing to complex projects over an extended period.

Compensation

Compensation is output-based. Experts are paid for tasks that meet project specifications and quality requirements. The time required for each task will vary based on its complexity and the expert’s workflow.