Personalized Product & Color Palette Recommender
Benefits
This agent delivers hyper-personalized product recommendations by analyzing customer purchase behavior, product interactions, and preference trends. It batches customers, evaluates their top categories, brands, and color palettes, and returns tailored product matches with images.
Problem Addressed
Retail brands struggle to offer timely, relevant, and visually aligned product recommendations that match both short-term behavior and long-term preferences. Manual methods can't scale or personalize deeply enough.
What the Agent Does
This AI agent performs batch-wise recommendations across customers:
• Get Dataset Row Count: Determines number of customers to process in each batch
• Batch Loading: Skips batch if no rows exist; continues if rows found
• Personalized Product Recommender:
o Extracts 1-year customer engagement (purchases + interactions)
o Identifies top category, brand, color palette, and style tags
o Considers long-term preferences (older than 1 year)
o Filters products from product_details table:
▪ In stock = TRUE
▪ Matches top color palettes, category, brand, and style tags
o Recommends up to 3 best matches per customer
• Append to Dataset: Stores final recommendations
• Add Numbers: Aggregates totals for downstream reporting
Standout Features
• Combines short-term and long-term behavior
• Color-aware personalization logic
• Returns image links for email embedding
• Fallback formatting logic for fewer than 3 matches
