Mandarin question answer pairs for electrical engineering

About OpenTrain

OpenTrain (opentrain.ai) aggregates data-labeling and AI-training jobs from many AI companies and labeling platforms into a single place so contributors can discover and apply quickly. Creating an OpenTrain account is free and applying takes only a few minutes.

We list roles like this one to connect skilled contributors with short-term, flexible labeling and review projects that power the next generation of AI systems.

About AI training and RAG work

AI training (also called data labeling, annotation, or human feedback work) is the human side of building AI: people create, check, and rate examples that models learn from. This posting focuses on Retrieval-Augmented Generation (RAG) workflows, where model outputs must be fully supported by cited source documents.

Your task will contribute to supervised fine-tuning and RAG datasets: ensuring answers are accurate, high-quality, and explicitly backed by the provided document set so downstream models can generate grounded responses.

The role

You will review 50–100 Mandarin question-and-answer pairs in electrical engineering, evaluate existing feedback (good/bad), and find matching supporting documents from a provided document pool using a document-finding tool. Every answer must be fully supported and cited by documents available to you.

Work is contractor, part-time, and remote (worldwide). This is an intermediate-level content review and citation task focused on technical accuracy and source matching.

• Label type: PROMPT_RESPONSE_WRITING_SFT (supervised fine-tuning style review).
• Data type: TEXT. Labeling software: OTHER (you will use our document-finder tool and provided workflow).

What you'll do

For each Q&A pair you will: review the question and answer in Mandarin, assess the existing feedback (good/bad), locate one or more documents from the provided corpus that directly support the answer, and record explicit citations or document IDs per our instructions.

Aim for clarity and precision: mark when answers are unsupported or partially supported, and attach the minimal set of documents that fully justify the answer. Use the document-finding tool supplied to search the corpus and record matching results in the required fields.

• Review and re-evaluate feedback labels (good/bad) for each pair.
• Find and attach supporting document(s) for every answer; cite document IDs or URLs as requested.
• Flag answers that cannot be fully supported by the provided documents.

Requirements

Applicants must be fluent in Mandarin (reading and writing) and able to assess technical material in Chinese. The subject matter is electrical engineering; familiarity with electrical engineering concepts is required.

This is an intermediate-level role. Ideally you have formal education or work experience in electrical engineering; experience at a large semiconductor company is a strong plus but not mandatory.

• Mandarin fluency (reading and writing) is required.
• Background in electrical engineering (education or employment) preferred.
• Experience at a large semiconductor company is a significant bonus.

Compensation, schedule, and logistics

Pay is hourly at USD 40. This is a contractor, part-time engagement. You should expect to work 20+ hours per week. Based on our estimates, the full workload will typically take about 20–40 hours overall depending on your speed and accuracy.

This project is fully remote and open worldwide. The workflow uses a document-finding tool plus the labeling interface we provide (software categorized as OTHER).

• Rate: $40 per hour (USD).
• Time: 20+ hours/week; project estimated at ~20–40 total hours depending on performance.
• Employment types: Contractor, Part-time. Remote work, worldwide.

Who should apply

You should apply if you are fluent in Mandarin, comfortable reading technical electrical-engineering material, and can commit to part-time contractor hours. Intermediate reviewers who can rapidly find and cite supporting documents in a RAG workflow will excel.

If you have industry experience at a semiconductor firm, or prior annotation/review work in technical domains, mention it in your application to help us match you to this dataset.

How the process works

After you’re hired you will receive access to the dataset of Q&A pairs, the document corpus, and the document-finding tool. Follow the provided instructions to attach document IDs/citations for each answer and to mark feedback labels.

OpenTrain makes applying quick: create a free account, complete your profile, and submit your interest. Our team will provide the onboarding checklist and access details once you’re selected.

• You will use a provided document-finding tool and the labeling workflow (software listed as OTHER).
• Every answer must be fully supported by the provided documents; cite documents exactly as requested.