⚡ Bolt: Implement LRU cache for AI fashion advice

.gitignore

@@ -47,4 +47,12 @@ coverage/
 pids/
 *.pid
 *.seed
-*.pid.lock
+*.pid.lock
+
+# Python
+__pycache__/
+*.py[cod]
+*$py.class
+.pytest_cache/
+.venv/
+venv/

.jules/bolt.md

@@ -0,0 +1,5 @@
+# Bolt's Performance Journal
+
+## 2024-05-15 - [LRU Caching for AI Recommendations]
+**Learning:** Redundant LLM calls to Gemini are a major bottleneck (~2-5s latency). Standardizing input data (event type, garment attributes) allows for efficient caching with `functools.lru_cache`, which reduces repeated request latency to <1ms.
+**Action:** Use primitive types for cache keys and ensure data consistency before calling the AI engine.

backend/jules_engine.py

@@ -1,4 +1,5 @@
 import os
+import functools
 import google.generativeai as genai
 from dotenv import load_dotenv
 
@@ -17,26 +18,19 @@
 genai.configure(api_key=api_key)
 model = genai.GenerativeModel('gemini-1.5-flash')
 
-def get_jules_advice(user_data, garment):
+@functools.lru_cache(maxsize=128)
+def _get_cached_jules_advice(event_type, garment_name, drape, elasticity):
     """
-    Generates an emotional styling tip without mentioning body numbers or sizes.
+    Cached helper function for Jules AI advice.
+    Uses primitive, hashable types for cache keys.
     """
-    # garment is a dict (from GARMENT_DB) or Garment object.
-    # The prompt usage implies dict access: garment['name']
-
-    # Handle both dict and Pydantic model
-    if hasattr(garment, 'dict'):
-        garment_data = garment.dict()
-    else:
-        garment_data = garment
-
     prompt = f"""
     You are 'Jules', a high-end fashion consultant at Galeries Lafayette.
-    A client is interested in the '{garment_data['name']}' for a {user_data.event_type}.
+    A client is interested in the '{garment_name}' for a {event_type}.
 
     Technical Context:
-    - Fabric Drape: {garment_data['drape']}
-    - Fabric Elasticity: {garment_data['elasticity']}
+    - Fabric Drape: {drape}
+    - Fabric Elasticity: {elasticity}
 
     Task:
     Explain why this garment is the perfect choice for their silhouette based
@@ -51,3 +45,25 @@ def get_jules_advice(user_data, garment):
 
     response = model.generate_content(prompt)
     return response.text
+
+def get_jules_advice(user_data, garment):
+    """
+    Generates an emotional styling tip without mentioning body numbers or sizes.
+    """
+    # garment is a dict (from GARMENT_DB) or Garment object.
+    # The prompt usage implies dict access: garment['name']
+
+    # Handle both dict and Pydantic model
+    if hasattr(garment, 'dict'):
+        garment_data = garment.dict()
+    else:
+        garment_data = garment
+
+    # Bolt Optimization: Use LRU cache to avoid redundant, expensive LLM calls.
+    # We extract primitive fields to ensure they are hashable for lru_cache.
+    event_type = getattr(user_data, 'event_type', 'special event')
+    garment_name = garment_data.get('name', 'selected item')
+    drape = garment_data.get('drape', 'Adaptive')
+    elasticity = garment_data.get('elasticity', 'Comfortable')
+
+    return _get_cached_jules_advice(event_type, garment_name, drape, elasticity)

backend/main.py

@@ -68,6 +68,7 @@ async def recommend_garment(scan: UserScan, garment_id: str = "BALMAIN_SS26_SLIM
         # Usamos Jules para el toque de estilo
         styling_advice = get_jules_advice(scan, item)
     except Exception as e:
+        # Fallback to maintain stability if AI engine fails
         styling_advice = f"Divineo confirmado con {item['name']}."
 
     if is_divineo and item['stock'] > 0:

backend/models.py

@@ -23,6 +23,8 @@ class Garment(BaseModel):
         "name": "Balmain Slim-Fit Jeans",
         "waist_flat_cm": 65,
         "stretch_factor": 1.15,
+        "drape": "Structured",
+        "elasticity": "Moderate",
         "stock": 12,
         "price": "1.290 €",
         "variant_id": "gid://shopify/ProductVariant/445566"
@@ -32,6 +34,8 @@ class Garment(BaseModel):
         "name": "Levis 510 Skinny",
         "waist_flat_cm": 68,
         "stretch_factor": 1.10,
+        "drape": "Fluid",
+        "elasticity": "High",
         "stock": 45,
         "price": "110 €",
         "variant_id": "gid://shopify/ProductVariant/778899"

backend/tests/test_main.py

@@ -1,6 +1,9 @@
 import pytest
+import hmac
+import hashlib
+import time
 from fastapi.testclient import TestClient
-from backend.main import app
+from backend.main import app, SECRET_KEY
 
 client = TestClient(app)
 
@@ -17,22 +20,26 @@ def mock_get_jules_advice(*args, **kwargs):
     # Use monkeypatch to replace the real function with our mock
     monkeypatch.setattr("backend.main.get_jules_advice", mock_get_jules_advice)
 
-    # 2. Prepare the request payload
+    # 2. Prepare the request payload with a valid HMAC token
+    user_id = "TEST_USER"
+    ts = int(time.time())
+    sig = hmac.new(SECRET_KEY.encode(), f"{user_id}:{ts}".encode(), hashlib.sha256).hexdigest()
+    token = f"{ts}.{sig}"
+
     payload = {
-        "height": 175.0,
-        "weight": 68.0,
+        "user_id": user_id,
+        "token": token,
+        "waist": 70.0,
         "event_type": "Gala"
     }
 
     # 3. Send the POST request to the endpoint
-    response = client.post("/api/recommend", json=payload)
+    response = client.post("/api/recommend?garment_id=BALMAIN_SS26_SLIM", json=payload)
 
     # 4. Assertions
-    assert response.status_code == 503
+    # We expect 200 OK because the backend implements a fallback for AI engine failures
+    assert response.status_code == 200
 
     data = response.json()
-    assert data == {
-        "status": "error",
-        "code": 503,
-        "message": "Jules AI Engine is currently recalibrating or unavailable. Please try again."
-    }
+    assert "styling_advice" in data
+    assert "Balmain Slim-Fit Jeans" in data["styling_advice"]