- Company Name
- SME Careers
- Job Title
- Hebrew Trust & Safety Data Trainer
- Job Description
-
**Job title**
Hebrew Trust & Safety Data Trainer
**Role Summary**
Hourly, fully remote contractor tasked with reviewing AI‑generated responses, generating safety‑focused evaluation content, and providing expert feedback to ensure output accuracy, safety, and clarity in both English and Hebrew.
**Expactations**
* Deliver consistent, high‑throughput annotations
* Maintain precise, documented evaluations across time zones
* Demonstrate cultural‑linguistic nuance and policy‑consistent judgment
* Exhibit emotional resilience while handling explicit or toxic content
**Key Responsibilities**
- Curate and label safety training examples (including adversarial/red‑team cases) in English and Hebrew that probe model behavior on hate/harassment, sexual content, self‑harm, violence, bias, illegal services, malicious activity/code, and misinformation.
- Review, score, and compare multiple model responses against safety policy and quality rubrics, documenting reasons for safe/unsafe outcomes and identifying failure modes (evasion, normalization, escalation, procedural enablement).
- Continuously stress‑test and audit model behavior for policy gaps and edge cases, flag ambiguous scenarios, propose clearer decision rules, and help maintain consistent annotation standards across reviewers.
- Provide clear, concise documentation of decisions and support the iterative improvement of safety training data.
**Required Skills**
- Near‑native or native Hebrew proficiency (reading/writing).
- Minimum C1 English proficiency (reading/writing).
- Proven experience in Trust & Safety, content moderation, policy enforcement, risk operations, or investigations.
- Mandatory LLM red‑teaming experience with documented ability to probe safety boundaries.
- Deep knowledge of safety domains: hate & harassment, sexual content, self‑harm, violence, bias, illegal goods/services, malicious activities, malicious code, and misinformation.
- Strong judgment under ambiguity, ability to apply written policies consistently, and concise decision explanations.
- Emotional resilience for handling explicit or toxic content.
- Independent contractor mindset: dependable throughput, clear documentation, and responsiveness across time zones.
- Proficiency with AI tools such as ChatGPT, Gemini, Perplexity, and annotation platforms.
**Required Education & Certifications**
- Bachelor’s degree or higher in Communications, Linguistics, Psychology, Law/Policy, Security Studies, or equivalent professional experience.
- Certifications in Trust & Safety or content moderation are a plus, but LLM red‑teaming proof is mandatory.