3. Use Cases - What MangaAssist Does

Use Case Map

mindmap
  root((MangaAssist Use Cases))
    Pre-Purchase
      Product Discovery
      Search and Recommendation
      Edition Comparison
      Pricing and Deals
    During Purchase
      Checkout Help
      Cart Questions
      Gift Wrapping
    Post-Purchase
      Order Tracking
      Returns and Refunds
      Delivery Issues
    Support
      FAQ Answering
      Human Escalation
      Abuse Reporting
    Growth
      Cross-sell / Upsell
      New Release Alerts
      Wishlist Suggestions

Detailed Use Cases

UC-1: Product Discovery

Trigger: User asks "What manga should I read?" or "Show me popular shonen manga."

What happens: 1. Chatbot identifies the intent as discovery. 2. If the user gives a genre, mood, or theme, the recommendation engine is queried with those signals. 3. If the user gives no specifics, the chatbot asks one clarifying question such as "What genres do you enjoy? Action, romance, horror, sci-fi?" 4. Results are returned as a curated list of 3 to 5 titles with cover images, ratings, and one-line descriptions. 5. Each result is clickable and links to the product detail page.

Systems involved: Recommendation Engine, Product Catalog, User Profile if logged in.

UC-2: Search and Recommendation

Trigger: "Something like Attack on Titan but with more political intrigue."

What happens: 1. The chatbot extracts the seed title (Attack on Titan) and the modifier (political intrigue). 2. It queries the recommendation engine with the seed ASIN and applies genre and tag filtering. 3. Optionally, RAG retrieves editorial descriptions or community tags that match "political intrigue." 4. Returns tailored recommendations such as Vinland Saga, Kingdom, and Legend of the Galactic Heroes.

Systems involved: Recommendation Engine, Product Catalog, RAG Knowledge Base.

UC-3: FAQ Support

Trigger: "What's the return policy for manga?" or "Do you price match?"

What happens: 1. Intent is classified as FAQ. 2. RAG retrieves the relevant policy document chunk from the JP Manga knowledge base. 3. The LLM generates a natural-language answer grounded in the retrieved policy text. 4. The response includes a link to the full policy page.

Example Q&A

UC-4: Order Tracking

Trigger: "Where is my order?" or "When will my Naruto box set arrive?"

What happens: 1. User must be authenticated. 2. Chatbot calls the Order Service API with the user's customer ID. 3. If the user has one recent manga order, it returns tracking status directly. 4. If multiple orders exist, the chatbot asks which one. 5. Returns order status, carrier, tracking number, and estimated delivery date.

Systems involved: Order Service, Shipping and Logistics.

UC-5: Returns and Refund Guidance

Trigger: "I want to return Volume 5, it arrived damaged."

What happens: 1. Chatbot identifies the order and item. 2. Checks return eligibility such as return window and item category. 3. If eligible, provides a step-by-step return flow or deep-links to the return initiation page. 4. If not eligible, explains why and offers alternatives such as contact support or replacement.

Systems involved: Order Service, Returns Service, Customer Profile.

UC-6: Promotion and Discount Awareness

Trigger: "Are there any deals on manga right now?" or user is browsing a product page during a sale.

What happens: 1. Chatbot queries the Promotions Service for active deals in the manga category. 2. Returns relevant deals such as "Buy 2, Get 1 Free on select shonen titles this week." 3. If the user is viewing a specific product, it checks whether that ASIN is part of any active promotion.

Systems involved: Promotions Service, Product Catalog.

UC-7: Personalized Suggestions

Trigger: Logged-in user opens the chatbot without a specific question.

What happens: 1. Chatbot greets the user and proactively offers suggestions. 2. Pulls recent browsing history, past purchases, and wishlist items. 3. Generates a personalized opening such as "Welcome back! Volume 12 of My Hero Academia just released - you bought Vol 11 last month. Want to add it to your cart?"

Systems involved: User Profile, Purchase History, Recommendation Engine, Product Catalog.

UC-8: Cross-sell and Upsell

Trigger: User adds a manga volume to cart, or asks about a specific title.

What happens: 1. Chatbot detects the product context from the current page ASIN or cart contents. 2. Suggests complementary products such as art books, box sets, or related series. 3. Suggests upgrades such as "The box set (Vols 1-11) is cheaper than buying individually."

Systems involved: Recommendation Engine, Product Catalog, Cart Service.

UC-9: Customer Support Escalation

Trigger: Chatbot cannot resolve the issue, or the user explicitly asks for a human.

What happens: 1. Chatbot acknowledges the limitation. 2. Packages the conversation summary, user context, and issue category. 3. Hands off to the Amazon Customer Service queue with full context so the user does not repeat themselves. 4. Provides estimated wait time.

Systems involved: Support Routing Service, Conversation Memory.

Use Case Priority Matrix