Content Evaluator - Bilingual (Vietnamese and English) - Flexible Hours

Innodata; United States; Hybride; USD 23 000 - 32 000; Plein temps

Résumé du poste

Innodata seeks detail-oriented evaluators to conduct human quality evaluations for an enterprise AI customer support product on Instagram, WhatsApp, and Messenger. You will evaluate and benchmark AI model responses against complex evaluation rubrics using provided business knowledge bases.

You’ll thrive if you have strong English and Vietnamese writing skills, a customer service background, and a disciplined approach to following multi-tier guidelines.

Qualifications

• Customer service experience in call centers, retail, or handling inquiries via email, chat, or phone.
• Excellent written English with strong tone, brand voice, grammar, and nuance.
• Excellent written Vietnamese with strong tone, brand voice, grammar, and nuance.
• Ability to strictly follow multi-tier evaluation guidelines and complex rubrics.

Responsabilités

• Review and score AI-generated customer interactions across dimensions like Action Fidelity, Faithfulness, Hallucination, Compliance, Tone, and Handoff.
• Benchmark informational (R1) and transactional (R2) queries against authoritative sources within the task UI.
• Participate in dual-review processes and calibration audits to ensure inter-rater agreement and ground-truth targets.
• Deliver precise evaluation within established performance targets.

Description du poste

Innodata(Nasdaq: INOD) is a global data engineering company. We believe that data and Artificial Intelligence (AI) are inextricably linked.Our mission is to enable the responsible advancement of artificial intelligence by providing the data, evaluation frameworks, and human expertise required to build AI systems that can be trusted at scale.We provide a range of transferable solutions, platforms, and services for Generative AI / AI builders and adopters. In every relationship, we honor our 36+ year legacy delivering the highest quality data and outstanding outcomes for our customers.

Role Overview

We are seeking detail-oriented evaluators to conduct human quality evaluations for an enterprise AI customer support product on Instagram, WhatsApp, and Messenger. You will evaluate and benchmark AI model responses against complex evaluation rubrics using provided business knowledge bases.

What You’ll Own:

• Model Evaluation: Review and score AI-generated customer interactions across Foundational, Experiential, and Operational dimensions (e.g., Action Fidelity, Faithfulness, Hallucination, Compliance, Tone, and Handoff).
• Intent & Fact Verification: Benchmark both informational (R1) and transactional (R2) customer queries against authoritative business sources (FAQs, product catalogs, SOPs) within the task UI.
• Quality Assurance: Participate in dual-review processes and daily calibration audits to ensure inter-rater agreement and establish ground-truth performance targets.
• Performance Targets: Deliver precise evaluation

You’ll Thrive in This Role If You Have:

• Customer Service Background: Prior experience in customer service, call centers, retail, or handling customer communications via email, chat, or phone (highly prioritized).
• English Proficiency: Exceptional written English skills with a strong command of tone, brand voice, grammar, and nuance.
• Vietnamese Proficiency: Exceptional written Vietnamese skills with a strong command of tone, brand voice, grammar, and nuance.
• Analytical Precision: Ability to strictly follow multi-tier evaluation guidelines, complex logic trees, and technical rubrics without deviation.
• Tech Adaptability: Comfort using dedicated web-based tools and labeling interfaces.

The expected hourly salary range for this position is up to $20 p/hour, based on experience, skills, and qualifications.