Independent AI research, prompt testing, and LLM response evaluation (quality analyst style)
Independently tested multiple LLMs to evaluate how accurately they answer complex engineering and mathematics questions. Performed quality checks on model responses by verifying logic, identifying hallucinations, and confirming factual correctness for school-related technical projects. Compared model performance across speed and accuracy to guide which model generated more reliable outputs. • Tested LLM responses for engineering/math question correctness • Checked responses for hallucinations and logical errors • Compared DeepSeek vs. Gemini on engineering problem solving • Wrote and iterated detailed prompts to improve output usefulness