Train, tune and distribute models with generative AI and machine learning capabilities
Step 1: Establish Assessment Framework
Assign the model an expert role in your target domain (e.g. "You are an expert UX designer").
Prompt it to analyze the original input and score its overall quality on a 1–100 scale.
Record this "Baseline Score."
Step 2: Set Improvement Threshold
Define a measurable uplift goal (e.g. Baseline + 15 – 20%).
Instruct the model to improve the subject until it meets or exceeds that threshold.
Ask it to produce a concrete artifact (diagram, revised text, prototype outline, etc.) that illustrates the enhanced concept.
Step 3: Initiate Meta-Analysis
Prompt the model to review the entire interaction and the enhancement technique used.
Have it evaluate the effectiveness of the workflow itself, scoring the process from 1–100.
Generate a concise process-analysis report detailing strengths, weaknesses, and lessons learned.