AI Hallucination Detection Specialist vs AI Model Robustness Tester

A AI Hallucination Detection Specialist Only

• LLM output evaluation and benchmarking (BLEU, ROUGE, faithfulness scores, custom metrics)
• Prompt engineering and adversarial prompt design for stress-testing models
• Retrieval-Augmented Generation (RAG) architecture understanding and grounding verification
• Natural Language Inference (NLI) and entailment-based fact verification
• Confidence calibration and uncertainty quantification for generative models
• Automated evaluation pipeline design using OpenAI Evals, promptfoo, or custom frameworks
• Domain-specific knowledge graph construction and claim decomposition
• Statistical analysis of hallucination rates across model versions, prompts, and domains