Shanghainese Speech Evaluator (Remote, Part-Time)

About OpenTrain and This Opportunity

OpenTrain is the #1 platform for building careers in AI training and data labeling. We connect people with projects that shape how AI systems understand and speak real languages.

This role hires directly through OpenTrain: you’ll work as a remote contractor evaluating AI-generated speech in Shanghainese to improve pronunciation, accent, and overall naturalness.

• Employment type: Contractor, Part-time
• Workload: Less than 20 hours/week
• Pay: $23 USD per hour

About AI Training Work

AI training (also called data labeling or human feedback work) is the human side of building intelligent systems. Contributors provide careful examples, judgments, and corrections so models learn real-world language and sound natural.

These projects are highly flexible, often remote, and accessible to people with language expertise — you’ll have direct impact on how a model speaks Shanghainese.

• 100% remote work — join from anywhere with a stable internet connection
• Flexible schedule — fit tasks around other commitments
• No special software required beyond your laptop/PC and internet

The Role

We are seeking linguists and fluent Shanghainese speakers to evaluate AI-generated speech. You will record audio prompts, listen to model responses, and rate pronunciation and fluency until high-quality, natural-sounding outputs are achieved.

Candidates should have strong knowledge of Shanghainese phonetics, accent variations, and fluency patterns. Prior experience in speech evaluation, linguistic analysis, or accent validation is preferred but not mandatory.

• Required: fluent Shanghainese knowledge; preferred: 2+ years of experience using Shanghainese in a linguistic or evaluative capacity
• Must have a personal laptop/PC and a stable internet connection
• Worldwide applicants accepted

What You'll Do

Your core task is to interact with an AI model by recording prompts and evaluating the model’s spoken responses for pronunciation and fluency. You will not evaluate the content of responses — only how natural and accurate the speech sounds.

You will iterate with the model: generate, listen, rate, and provide structured feedback until at least two (and up to five) responses receive a 7/7 rating for naturalness. Some prompts may require many attempts (10, 20, or more) to reach the target.

• Record clear audio prompts in Shanghainese and submit model responses for review
• Rate audio using an evaluation rubric (aiming for 7/7 naturalness on multiple outputs)
• Provide short, actionable feedback focused on pronunciation, intonation, and fluency
• Repeat iterations to refine model output as needed

Requirements

We want reliable evaluators who can give consistent, objective ratings and clear corrective guidance. Preserve accuracy and be able to explain phonetic or accent-specific issues concisely.

Note: the project is labeled as entry level but prefers candidates with practical experience in Shanghainese; include any relevant background when you apply.

• Fluency in Shanghainese and strong knowledge of its phonetics and accent variation
• Access to a laptop or PC and stable internet
• Ability to record clear audio and review multiple iterations
• Preferred: 2+ years’ experience with Shanghainese or experience in speech evaluation/linguistic analysis

Who Should Apply

Apply if you are a native or near-native Shanghainese speaker who enjoys detailed listening work and can give structured, actionable feedback on spoken language.

This role suits linguists, language teachers, advanced bilingual speakers, or anyone with interest and skill in accent and fluency assessment.

• Good fit for people seeking flexible, part-time remote work related to language and speech
• Great opportunity to influence how AI systems speak Shanghainese