Business Analyst
TextEvaluation Rating
- Evaluated LLM-generated responses for accuracy, coherence,
helpfulness, and safety across multimodal inputs
(text, image, audio, video)
- Designed and engineered prompts to test model behavior
across edge cases and real-world scenarios
- Annotated and labeled 1000+ data samples across text,
image, and multimodal datasets for AI training pipelines
- Assessed model outputs using RLHF principles — ranking,
rating, and providing structured human feedback
- Identified failure patterns, hallucinations, and biases
in LLM responses and documented reproducible error cases
- Contributed to quality control workflows ensuring
consistency across annotation teams
- Evaluated LLM-generated responses for accuracy, coherence, helpfulness, and safety across multimodal inputs (text, image, audio, video) - Designed and engineered prompts to test model behavior across edge cases and real-world scenarios - Annotated and labeled 1000+ data samples across text, image, and multimodal datasets for AI training pipelines - Assessed model outputs using RLHF principles — ranking, rating, and providing structured human feedback - Identified failure patterns, hallucinations, and biases in LLM responses and documented reproducible error cases - Contributed to quality control workflows ensuring consistency across annotation teams
2026 - Present