pylibseekdb-runtime 1.0.0.post1

OceanBase SeekDB

🔷 The AI-native hybrid search database

Powerful AI search capabilities · Lightweight · Production-ready

---

---

🚀 What is OceanBase SeekDB?

OceanBase SeekDB is the lightweight, embedded version of OceanBase Database - a powerful AI search database designed for the AI applications. It combines enterprise-grade database capabilities with cutting-edge **AI search ** features, such as Vector search, fulltext search, Json.

---

🔥 Why OceanBase SeekDB?

---

Feature	OceanBase SeekDB	Traditional DB	Vector-only DB	Full-Text Engine
Embedded Mode	✅ Native	⚠️ Possible	⚠️ Possible	⚠️ Possible
SQL Support	✅ Full SQL	✅ Full SQL	❌ Limited	❌ Limited
Vector Search	✅ Built-in	❌ Limited	✅ Supported	❌ Limited
Full-Text Search	✅ Built-in	✅ Supported	❌ Limited	✅ Advanced
Json	✅ Yes	⚠️ Varies by Product	❌ Not Supported	❌ Not Supported
ACID Transactions	✅ Full	✅ Full	❌ Limited	❌ Limited
Easy Migration	✅ MySQL Compatible	✅ Standard	❌ No	❌ No

---

✨ Key Features

🎯 AI-Powered Search

• Vector Similarity Search: Optimize vector query accuracy, performance, and cost for different scenarios using different algorithms
• Hybrid Search: Combine vector search, scalar search, and full-text retrieval for optimal results
• Full-Text Search: Built-in full-text indexing for keyword-based searches
• Json：Built-in Json schema and query, support Json index.

📦 Embedded & Lightweight

• Zero Dependencies: Run embedded in your application - no separate database server required
• Tiny Footprint: Minimal memory and disk usage, perfect for edge devices and containers
• Single Binary: Easy to deploy and distribute with your application
• Local-First: Work offline, sync when ready

⚡ Simple & Developer-Friendly

• MySQL Compatible: Use familiar SQL syntax - no learning curve
• Instant Setup: Get started in seconds, not minutes
• Rich APIs: Support for Python, Java, Go, and more
• Comprehensive Docs: Clear documentation with examples for every use case

🚀 Production-Ready

• Stability: More than 15 years of technical expertise and 4000+ enterprise implementations
• ACID Compliance: Full transaction support with strong consistency guarantees
• Horizontal Scalability: Scale from single node to distributed cluster seamlessly
• Enterprise Security: Built-in encryption, authentication, and access control

---

🎬 Quick Start

Installation

pip install pylibseekdb

Requirements

• CPython 3.8+
• Linux x86_64, aarch64/arm64 with glibc version >= 2.28 (Alpine Linux is not supported yet)

🎯 AI Search Example

Build a semantic search system in 5 minutes:

🗄️ 🐍 Python SDK

# install sdk first
pip install -U pyseekdb

import pyseekdb
client = pyseekdb.Client()

collection = client.get_or_create_collection(name="my_collection")

collection.upsert(
    ids=["1", "2", "3"],
    documents=[
        "It's rainy today",
        "It was cloudy yesterday",
        "The forecast for tomorrow is fine weather"
   ]
)

results = collection.query(
    query_texts=[" What's the weather like today"], # SeekDB will embed this for you
    n_results=2 # how many results to return
)

print(results)

🗄️ SQL

import pylibseekdb as seekdb

# Open a database
seekdb.open()

# Connect to a database
conn = seekdb.connect()

# Use the connection
cursor = conn.cursor()
cursor.execute("""-- Create table with vector column
CREATE TABLE articles (
    id INT PRIMARY KEY,
    title TEXT,
    content TEXT,
    embedding VECTOR(384)
);""")

cursor.execute("""-- Create vector index for fast similarity search
CREATE INDEX idx_vector ON articles USING VECTOR (embedding);""")

cursor.execute("""-- Insert documents with embeddings
-- Note: Embeddings should be pre-computed using your embedding model
INSERT INTO articles (id, title, content, embedding) 
VALUES 
    (1, 'AI and Machine Learning', 'Artificial intelligence is transforming...', '[0.1, 0.2, ...]'),
    (2, 'Database Systems', 'Modern databases provide high performance...', '[0.3, 0.4, ...]'),
    (3, 'Vector Search', 'Vector databases enable semantic search...', '[0.5, 0.6, ...]');""")

cursor.execute("""-- Example: Hybrid search combining vector and full-text
-- Replace '[query_embedding]' with your actual query embedding vector
SELECT 
    title,
    content,
    embedding <-> '[query_embedding]' AS vector_distance,
    MATCH(content) AGAINST('your keywords' IN NATURAL LANGUAGE MODE) AS text_score
FROM articles
WHERE MATCH(content) AGAINST('your keywords' IN NATURAL LANGUAGE MODE)
ORDER BY vector_distance ASC, text_score DESC
LIMIT 10;""")

results = cursor.fetchall()

# Close the connection
conn.close()

📚 Use Cases

🎯 RAG Systems	🔍 Semantic Search	💬 Chatbots	🎬 Recommendations
Build retrieval-augmented generation pipelines with vector search	Power semantic search across documents, images, and multimedia	Create intelligent chatbots with memory and context	Build recommendation engines with hybrid search

🎯 Real-World Examples

• 📚 Document Q&A: Build ChatGPT-like document search with RAG
• 🖼️ Image Search: Find similar images using vision embeddings
• 💼 E-commerce: Semantic product search and recommendations
• 🔬 Scientific Research: Search through research papers and datasets
• 📊 Business Intelligence: Combine SQL analytics with AI search

---

🌟 Ecosystem & Integrations

---

🤝 Community & Support

---

📄 License

OceanBase SeekDB is licensed under the Apache License, Version 2.0.