Project Hedgehog (Handshake AI) — AI model training and evaluation with labeled data and feedback
Trained and evaluated AI models by providing high-quality labeled data and structured feedback to support prompt refinements. Assessed model outputs for accuracy, reasoning quality, and consistency with the project’s guidelines. Worked with a distributed team to improve overall AI performance for real-world scenarios. • Labeled training/evaluation data according to project requirements. • Provided feedback and revised prompts to improve response quality. • Verified outputs against accuracy and alignment criteria. • Collaborated with teammates to iteratively enhance model behavior.