SWE Bench Data Engineer / Data Scientist — Benchmark Evaluation

About OpenTrain

OpenTrain is the centralized, open platform where people build careers in AI training and data labeling. Contributors use OpenTrain to find projects, consolidate work history, and develop a reusable portfolio that demonstrates their skills across AI evaluation and annotation work.

OpenTrain AI is the hiring and contracting organization for this role. Creating an OpenTrain account is free and lets you apply, manage opportunities, and build a record of benchmark and annotation work that employers respect.

• A focused career platform for people who train and evaluate AI systems
• Work is remote, flexible, and directly shapes how modern AI behaves
• Free to join and build a unified portfolio across projects

About AI Training Work

AI training (data labeling, annotation, and human evaluation) is the human work behind modern AI systems. Contributors annotate, validate, and review data and model outputs so researchers and engineers can measure and improve model behavior.

This role sits at the intersection of data engineering and data science: you will support benchmark-driven evaluations that mimic production workflows to ensure results are correct, reproducible, and useful for model development.

• 100% remote, often flexible hours and part-time-friendly
• Work ranges from beginner-friendly annotation to advanced engineering and evaluation
• Contributors directly affect how state-of-the-art AI systems are assessed

The Role

We are hiring experienced SWE Bench Data Engineers / Data Scientists to perform benchmark-driven evaluation work focused on real-world data engineering and data science workflows. The work centers on structured and unstructured datasets, production-like data pipelines, feature preparation, data validation, and Python-based experimentation in complex codebases.

This is a contract, part-time role expecting 20+ hours per week. Candidates must be fluent in spoken and written English and be located in one of the approved countries: IN, PK, NG, KE, EG, GH, BD, TR, BR, MX. Experience level: Intermediate (3+ years).

• Employment types: Contractor, Part-time
• Time requirement: 20+ hours/week
• Languages: English (spoken and written)
• Eligible countries: IN, PK, NG, KE, EG, GH, BD, TR, BR, MX

What You'll Do

• Build, validate, and maintain data pipelines for benchmarking and evaluation workflows
• Process, analyze, and prepare structured and unstructured data for data-science use cases
• Write, run, and modify Python code to support local experiments and data processing
• Review transformations and outputs for correctness, quality, and reproducibility
• Produce clean, reusable, and well-documented workflows suited to benchmark tasks
• Collaborate on code reviews and task design for real-world AI evaluation work

Requirements

You must have practical, demonstrable experience that matches the skills below. We will evaluate your ability to work with real codebases and produce reproducible data workflows.

• 3+ years as a data engineer, data scientist, or data-focused software engineer
• Strong Python ability for data processing, analysis, and workflow validation
• Experience with structured and unstructured data in real-world codebases
• Solid machine learning and data-science fundamentals
• Ability to navigate and modify complex, production-like codebases
• Ability to review code for maintainability, correctness, and reproducibility
• Excellent spoken and written English communication skills