Last Updated on September 8, 2026 by Justin Bryant
Innodata is accepting U.S.-based remote applications for a Content Evaluator – Bilingual (Vietnamese and English) – Flexible Hours. The job involves reviewing AI-generated customer-support responses on Instagram, WhatsApp, and Messenger, with listed pay of up to $20 per hour.
The key limitation is language fluency: applicants need exceptional written Vietnamese and English, including grammar, tone, brand voice, and nuance. Innodata also says prior customer service, call-center, retail, or customer-communication experience is highly prioritized.
The official application was active when checked on September 8, 2026. Greenhouse job ID: 4384507009.
What Is Innodata?
Innodata is a publicly traded data-engineering company that supports artificial-intelligence developers and adopters. Its work includes data preparation, model evaluation, quality systems, and human review used to improve AI products.
This opening supports an enterprise AI customer-service product. Evaluators do not simply rate whether an answer sounds good. They compare model responses with official business information and detailed scoring rules to judge accuracy, compliance, tone, and whether the AI took the proper action.
Innodata includes a recruiting-scam warning in the listing and says it will never ask applicants for payment, banking details, or sensitive financial information during the application process.
Pay and Benefits
The official posting lists an expected pay range of up to $20 per hour, based on the applicant's experience, skills, and qualifications. It does not provide a minimum rate or guarantee that every selected evaluator will receive $20 per hour.
The listing does not state whether this is W-2 employment, independent contracting, temporary employment, or project-based contributor work. It also does not list health insurance, paid time off, retirement benefits, bonuses, paid training, expense reimbursement, or guaranteed weekly hours. Applicants should confirm classification, minimum pay, expected workload, payment schedule, and benefit eligibility before accepting an offer.
Job Highlight: Bilingual Content Evaluator
The evaluator reviews AI-generated conversations intended for customer service on Meta messaging platforms. Using business knowledge bases, FAQs, product catalogs, standard operating procedures, and evaluation rubrics, the worker decides whether each model response is accurate and follows the required process.
The work can include checking factual faithfulness, identifying hallucinations, judging tone and compliance, confirming that an AI assistant completed the expected action, and deciding whether a conversation should have been handed off to a human. Evaluators also participate in second-review and calibration processes so multiple reviewers apply the rules consistently.
This is closer to structured content-quality review and AI evaluation than ordinary data entry. The tasks require careful reading, strong bilingual writing judgment, and the patience to follow complex instructions without substituting personal preference for the rubric.
Key Responsibilities
- Review and score AI-generated customer interactions.
- Evaluate accuracy, action fidelity, tone, compliance, and handoff decisions.
- Compare responses with official FAQs, catalogs, procedures, and business records.
- Check informational and transactional customer queries for factual correctness.
- Use web-based task and labeling interfaces.
- Follow detailed evaluation rubrics and logic trees consistently.
- Participate in dual-review quality checks.
- Join calibration audits intended to improve reviewer agreement.
- Help establish reliable ground-truth performance targets for the AI product.
- Document judgments precisely enough for quality review.
Qualifications
The official listing emphasizes:
- Exceptional written English, including grammar, tone, brand voice, and nuance.
- Exceptional written Vietnamese with the same level of language judgment.
- An ability to follow multi-level evaluation rules and complex logic precisely.
- Comfort using dedicated web-based tools and labeling interfaces.
- Strong attention to detail and analytical consistency.
Previous experience in customer service, call centers, retail, email, chat, phone support, or other customer communications is described as highly prioritized. The posting does not state a required number of years or list a college degree as a job requirement, although the application asks candidates to report their education.
Do not describe this as a no-experience job. Language expertise and the ability to judge customer-service conversations are central to the role, and related experience may materially affect selection.
Resume Keywords
Use these phrases only when they truthfully reflect your background:
- AI model evaluation
- Content evaluation
- Human feedback
- Quality assurance
- Data labeling
- Customer support
- Email support
- Chat support
- Bilingual Vietnamese
- Written English
- Tone and brand voice
- Fact verification
- Hallucination detection
- Action fidelity
- Compliance review
- Evaluation rubrics
- Knowledge bases
- Standard operating procedures
- Inter-rater agreement
- Calibration
- Ground truth
- Analytical precision
- Web-based labeling tools
- Messenger
Useful application evidence could include bilingual writing or editing work, customer-support quality scores, moderation experience, rubric-based review, accuracy improvements, or examples of catching factual and policy errors. Do not disclose private customer conversations or confidential business data.
Location and Schedule
The job is listed as Remote – United States, and the application asks which U.S. state the candidate lives in. The official posting does not identify excluded states, so applicants should confirm whether Innodata can engage or employ them from their location.
The title says Flexible Hours, but the listing does not define how workers choose their schedule, how many hours are available, whether there are minimum or maximum weekly hours, or whether work is continuous. It also does not state a time-zone requirement, start date, end date, weekend expectation, training schedule, or equipment policy.
The duties are performed through web-based tools and focus on written content. However, the posting does not promise zero meetings or phone communication, so it should not be marketed as a guaranteed no-phone job.
Closing Notes
This opening may fit a fluent Vietnamese-English writer with customer-service knowledge, careful reading habits, and an interest in AI quality evaluation. Its main advantages are remote U.S. eligibility, flexible-hours positioning, written content-review work, and a clear maximum hourly rate.
The biggest limitations are the bilingual requirement, preference for customer-communication experience, undisclosed minimum pay, and missing details about classification, duration, weekly hours, benefits, equipment, and state eligibility. The official posting does not list a closing date.
The official Greenhouse application was last verified as active on September 8, 2026. Job ID: 4384507009.
Note: If the “Apply Here” link no longer works, the job may have been filled. Subscribe below to be notified of future remote jobs.