Now you'll implement the two most sophisticated endpoints that demonstrate advanced Weaviate capabilities: targeted vector search and retrieval-augmented generation (RAG).

Endpoint 4: /explore - Genre-Based Discovery

Purpose: Discovers movies by genre with optional year filtering, sorted by popularity.

Learning focus: Use vector search as a first-pass, then sort results by object properties.

Your Task

Implement genre exploration with ranking:

@app.get("/explore", response_model=ExplorerResponse)
def explore_movies(
    genre: str = Query(..., description="Movie genre to explore"),
    year_min: Optional[int] = Query(
        None, description="Filter by release year - from this year"
    ),
    year_max: Optional[int] = Query(
        None, description="Filter by release year - to this year"
    ),
):
    """
    Explore movies by genre(s) and optional year
    - Returns most popular movies best matching the specified genre
    - Can filter by year
    - Sorted by popularity/rating
    """

Implementation Steps

1. Build year filters using the same logic as the search endpoint
2. Perform hybrid search targeting the "genres" vector specifically
3. Sort results by popularity score in descending order
4. Return organized results with the query parameters

Implementation Hints

• Use the appropriate target vector
• Build filters the same way as in /search: handle min only, max only, both, or neither
• You can use Python's sorted() function, with this lambda x: x["popularity"] as key
• Check that the sorting is done in descending order (most popular first)

Endpoint 5: /recommend - AI-Powered Recommendations

Purpose: Takes a viewing occasion (like "date night" or "family movie") and uses RAG to recommend suitable movies with explanations.

Learning focus: Combining retrieval with generative AI to create context-aware recommendations.

Your Task

Implement RAG-powered recommendations:

@app.get("/recommend", response_model=RecommendationResponse)
def recommend_movie(
    occasion: str = Query(
        ..., description="Viewing occasion (e.g., 'date night', 'family movie')"
    )
):
    """
    Get movie recommendations based on viewing occasion
    - Generates a query string from occasion
    - Performs semantic search against movie descriptions
    - Returns best match with reasoning
    """

Implementation Steps

1. Convert the occasion to a search query using the provided movie_occasion_to_query() helper function
2. Build the task prompt for the AI model using the provided template
3. Perform RAG with semantic search using near_text()
4. Configure the generative model at runtime

Implementation Hints

The prompt template full_task_prompt is already constructed for you.

For the Weaviate operation:

• Do you remember how to perform generative searches?
  • Use grouped task
• Set target_vector="default" for general semantic search
• Configure with claude-3-5-haiku-latest model

Testing Your Endpoints

Once you are finished, you can start the FastAPI server with:

python main.py

Then, you can navigate to http://localhost:8000/docs to explore the API documentation and test your endpoints interactively.

What's next?

Time to see these advanced patterns in action! Let's explore the complete solutions for genre discovery and RAG-powered movie recommendations.