MoteDB Documentation

The world's first AI-native embedded database designed for embodied intelligence scenarios.

It is an embedded database purpose-built for edge devices such as home robots, AR glasses, and industrial robotic arms, with native support for unified storage and querying of vectors, text, time series, and spatial coordinates. MoteDB treats multimodal data types as first-class citizens, providing low-latency online retrieval, strongly consistent data semantics, and query extensions tailored for embodied intelligence perception and decision-making.

📚 Documentation Index

Getting Started

• Quick Start Guide: ./01-quick-start.md - Get started with MoteDB in 5 minutes
• Installation & Configuration: ./02-installation.md - Installation, configuration, and deployment parameters

Core Features

• SQL Operations: ./03-sql-operations.md - Complete SQL syntax and query templates
• Batch Operations: ./04-batch-operations.md - High-performance batch writes (10-20x improvement)
• Transaction Management: ./05-transactions.md - MVCC, WAL, and savepoint usage
• API Reference: ./14-api-reference.md - All public APIs with examples

Index System

• Index Overview: ./06-indexes-overview.md - How the five index types work together
• Column Index: ./07-column-index.md - Equality and range queries
• Vector Index: ./08-vector-index.md - FreshDiskANN & rerank
• Full-Text Index: ./09-text-index.md - BM25 and tokenization plugins
• Spatial Index: ./10-spatial-index.md - R-Tree and geospatial queries
• Time Series Index: ./11-timestamp-index.md - Range scans and compression

Advanced Topics

• Performance Tuning: ./12-performance.md - Configuration, tuning, and monitoring
• Data Types: ./13-data-types.md - Value enum and schema design

Best Practices

• Production Experience: ./15-best-practices.md - Architecture, writes, and indexing strategies
• FAQ: ./16-faq.md - Debugging, deployment, and troubleshooting

🚀 Core Features

1. SQL Engine

Full SQL support, including subqueries, aggregation, JOINs, and index management.

let db = Database::open("data.mote")?;
db.execute("CREATE TABLE users (id INT, name TEXT, email TEXT)")?;
db.execute("INSERT INTO users VALUES (1, 'Alice', 'alice@example.com')")?;
let results = db.query("SELECT * FROM users WHERE id = 1")?;

2. Multimodal Indexes

Five index types supporting different data scenarios:

Index Type	Use Case	Performance Boost
Column Index (COLUMN)	Equality/range queries	40x
Vector Index (VECTOR)	KNN similarity search	100x
Full-Text Index (TEXT)	BM25 text search	50x
Spatial Index (SPATIAL)	Geospatial queries	30x
Time Series (TIMESTAMP)	Time range queries	20x

3. High-Performance Batch Operations

Batch inserts are 10-20x faster than row-by-row inserts:

// Batch insert 10,000 records
let mut rows = Vec::new();
for i in 0..10000 {
    let mut row = HashMap::new();
    row.insert("id".to_string(), Value::Integer(i));
    row.insert("name".to_string(), Value::Text(format!("User{}", i)));
    rows.push(row);
}

let row_ids = db.batch_insert_map("users", rows)?;
// Throughput: 737,112 rows/sec

4. MVCC Transactions

Full transaction support, including savepoints:

let tx_id = db.begin_transaction()?;

db.execute("INSERT INTO users VALUES (1, 'Alice', 25)")?;
db.savepoint(tx_id, "sp1")?;

db.execute("INSERT INTO users VALUES (2, 'Bob', 30)")?;
db.rollback_to_savepoint(tx_id, "sp1")?; // Only rolls back Bob

db.commit_transaction(tx_id)?;

🎯 Use Cases

• Embedded AI Applications: Robots, edge computing devices
• Vector Database: RAG, semantic search, recommendation systems
• Spatiotemporal Data: Geospatial locations, sensor data
• Full-Text Search: Document retrieval, log analysis
• Real-Time Analytics: Time series data

📊 Performance Metrics

• Batch Insert: 737,112 rows/sec (10,000 records)
• Vector Search: Latency < 5ms (95% recall)
• Column Index Query: 40x improvement
• Memory Usage: Core data structures < 100MB
• Transaction Throughput: 10,000 TPS

💡 Recommended Usage

1. Use the SQL API primarily - concise, powerful, and easy to use
2. Prefer batch operations - use batch_insert_map() instead of row-by-row inserts
3. Use indexes wisely - choose the appropriate index type based on your query patterns
4. Enable transactions - ensure data consistency

🔗 Quick Links

• GitHub Repository
• API Documentation

📝 Version Information

Documentation version: v0.1.0 Last updated: 2026-01-11

---

Next: Read the Quick Start Guide to begin using MoteDB