AI Product Recommendation

Human → AI

AI analyzes customer browsing and purchase history to generate personalized recommendations.

5 nodes · 5 edgesretail
agenthumandb
Visual
Collect Customer Datadb

Pull browsing history, purchase records, and preference signals.

sequentialAnalyze Preferences
Analyze Preferencesagent

AI identifies patterns, segments customer, and scores product affinity.

sequentialGenerate Recommendations
Generate Recommendationssystem

Produce ranked list of products with confidence scores.

sequentialMerchandiser Review
Merchandiser Reviewhuman

Human reviews recommendations for brand alignment and inventory.

conditionalPush to Storefront
fallbackAnalyze Preferences
Push to Storefrontapi

Deploy approved recommendations to product pages and email campaigns.

uc-product-recommendation.osop.yaml
osop_version: "1.0"
id: "ai-product-recommendation"
name: "AI Product Recommendation"
description: "AI analyzes customer browsing and purchase history to generate personalized recommendations."

nodes:
  - id: "browse_data"
    type: "db"
    subtype: "query"
    name: "Collect Customer Data"
    description: "Pull browsing history, purchase records, and preference signals."

  - id: "ai_analyze"
    type: "agent"
    subtype: "llm"
    name: "Analyze Preferences"
    description: "AI identifies patterns, segments customer, and scores product affinity."
    security:
      risk_level: "low"

  - id: "generate_recs"
    type: "system"
    name: "Generate Recommendations"
    description: "Produce ranked list of products with confidence scores."

  - id: "merchandiser_review"
    type: "human"
    subtype: "review"
    name: "Merchandiser Review"
    description: "Human reviews recommendations for brand alignment and inventory."
    security:
      approval_gate: true

  - id: "push_storefront"
    type: "api"
    subtype: "rest"
    name: "Push to Storefront"
    description: "Deploy approved recommendations to product pages and email campaigns."

edges:
  - from: "browse_data"
    to: "ai_analyze"
    mode: "sequential"
  - from: "ai_analyze"
    to: "generate_recs"
    mode: "sequential"
  - from: "generate_recs"
    to: "merchandiser_review"
    mode: "sequential"
  - from: "merchandiser_review"
    to: "push_storefront"
    mode: "conditional"
    when: "review.approved == true"
  - from: "merchandiser_review"
    to: "ai_analyze"
    mode: "fallback"
    label: "Adjust recommendation criteria"