We Want Your Expertise
Horizon is the talent network for LILT’s proprietary AI evaluation platform, Applied AI, designed to bring structured human expertise into every stage of the model lifecycle.
The 5 Styles of AI Assessment
The five assessment styles are how we approach assessing AI performance. Utilization of these various approaches is dependent on the nature of the project.
Conversational Q&A
Evaluate multi-turn responses for accuracy, reasoning, safety, and consistency across languages.
Model Validation
Ensure model outputs meet domain standards for quality, safety, and deployment readiness.
Classifications
Assess how accurately models categorize and interpret structured and unstructured content.
Comparison / Ranking
Compare outputs to identify the highest-quality response based on defined criteria.
Model Breaking
Stress-test models to uncover failure modes, hallucinations, and safety
How to Apply
Submit your application with your information and relevant credentials.
After review, selected candidates will receive two test tasks representing two of the five assessment styles.
Complete your assigned assessment tasks and submit them to our AI Data recruitment team.
