Semantic Search - Vector Panda Use Cases

See the Difference

Search by meaning, not keywords

Optimizing Web Performance 98% match

Improve page load times by implementing lazy loading, compressing images, minifying CSS/JS, and leveraging browser caching...

📄 Documentation 🕒 Updated 2 days ago

CDN Configuration Guide 94% match

Speed up content delivery by distributing static assets across global edge servers, reducing latency for users worldwide...

🔧 Tutorial 🕒 Updated 1 week ago

Database Query Optimization 89% match

Reduce response times by optimizing database queries, adding proper indexes, and implementing efficient caching strategies...

💡 Best Practice 🕒 Updated 3 days ago

Implementation Guide

Build semantic search in minutes

1. Index Your Content

Convert your documents, products, or any searchable content into embeddings and store them in Vector Panda with rich metadata.

Pythonimport veep
from sentence_transformers import SentenceTransformer
# Initialize
panda = veep.Client("your-api-key")
model = SentenceTransformer('all-MiniLM-L6-v2')
collection = panda.collection("products")
# Index your content
for product in products:
    text = f"{product.title} {product.description}"
    embedding = model.encode(text)
    
    collection.upsert(
        id=product.id,
        vector=embedding,
        metadata={
            "title": product.title,
            "description": product.description,
            "category": product.category,
            "price": product.price,
            "url": product.url
        }
    )

2. Build Search API

Create a search endpoint that converts queries to embeddings and retrieves semantically similar results from Vector Panda.

Pythonfrom fastapi import FastAPI
from typing import List
app = FastAPI()
@app.get("/search")
async def search(
    query: str,
    limit: int = 10,
    category: str = None
):
    # Convert query to embedding
    query_embedding = model.encode(query)
    
    # Search with optional filters
    filters = {}
    if category:
        filters["category"] = category
    
    results = collection.search(
        vector=query_embedding,
        k=limit,
        filters=filters,
        include_metadata=True
    )
    
    # Format results
    return [{
        "id": r.id,
        "title": r.metadata["title"],
        "description": r.metadata["description"],
        "score": r.score,
        "url": r.metadata["url"]
    } for r in results]

3. Enhance with Hybrid Search

Combine semantic search with traditional filters and facets for the best of both worlds. Vector Panda's metadata queries make this simple.

Python# Advanced search with multiple criteria
def hybrid_search(
    query: str,
    price_range: tuple = None,
    categories: List[str] = None,
    must_have_keywords: List[str] = None
):
    # Semantic search
    embedding = model.encode(query)
    
    # Build filters
    filters = {}
    if price_range:
        filters["price"] = {
            "$gte": price_range[0],
            "$lte": price_range[1]
        }
    if categories:
        filters["category"] = {"$in": categories}
    
    # Search with filters
    results = collection.search(
        vector=embedding,
        k=50,  # Get more results for re-ranking
        filters=filters,
        include_metadata=True
    )
    
    # Re-rank by keyword presence if needed
    if must_have_keywords:
        results = [r for r in results
                  if any(kw in r.metadata["title"].lower()
                       for kw in must_have_keywords)]
    
    return results[:10]

Traditional vs Semantic Search

Why semantic search delivers better results

Traditional Keyword Search

❌ Requires exact keyword matches
❌ Misses synonyms and related concepts
❌ Poor with typos and misspellings
❌ Language-specific implementations
❌ Complex query syntax for users

Vector Panda Semantic Search

✅ Understands meaning and intent
✅ Finds synonyms automatically
✅ Handles typos gracefully
✅ Multi-language support built-in
✅ Natural language queries

Semantic Search Features

Everything you need for modern search

🎯

100% Recall

Our PCA indexing guarantees you'll never miss relevant results. Perfect recall means happier users who find what they need.

🌍

Multi-Language

Search across languages seamlessly. A query in English can find results in Spanish, French, or any language your model supports.

🏷️

Metadata Filtering

Combine semantic search with traditional filters. Search by meaning while filtering by price, category, date, or any metadata.

⚡

Real-Time Updates

Add, update, or remove items instantly. No re-indexing needed. Your search results are always up-to-date.

📊

Relevance Scoring

Get similarity scores for each result. Fine-tune thresholds and implement custom ranking logic based on your needs.

🚀

Infinite Scale

From thousands to billions of items. Vector Panda scales automatically without configuration or performance degradation.

Natural Language Semantic Search