Applied Generative AI Practices

This unit advances learners’ practical and technical skills in generative AI by emphasizing prompt engineering, multimodal content creation, model behavior analysis, and system integration. Students will practice troubleshooting, code translation, evaluation techniques, and responsible-AI safeguards applied to creative and financial use cases to prepare for prototype deployment in the next unit.

Learning Objectives

• Analyze strengths, limitations, and failure modes of generative models across text, image, audio, and code modalities by comparing outputs against defined quality, fidelity, and alignment criteria
• Apply advanced prompt-engineering and human-in-the-loop training techniques (including reward/punishment, few-shot examples, data formatting, and citation integration) to iteratively improve creative and domain-specific outputs
• Demonstrate the ability to integrate and troubleshoot generative AI components within workflows by performing tasks such as speech-to-text integration, code translation (Matlab to Python), handling freezes/errors, and prototyping autonomous agent behaviors
• Evaluate ethical, legal, and safety risks (e.g., hallucination, bias, copyright, privacy) and develop a responsible-AI checklist and incident-response mitigation plan for a generative-AI project