driftdb-ai 0.1.4

The AI-native database client — graph memory + vector search + real-time sync in 5.3 MB

DriftDB Python Client

The AI-native database client. Graph memory + vector search + real-time sync in 5.3 MB.

pip install driftdb

Quick Start

from driftdb import DriftDB

db = DriftDB("http://localhost:9211", token="your-token")

# Create nodes
db.create_node(labels=["User"], properties={"name": "Amrit", "age": 25})

# Query with DriftQL
result = db.query('FIND (u:User) RETURN u.name, u.age')

# AI Memory — one line to give your LLM a brain
db.remember("User prefers dark mode")
db.remember("Meeting at 3pm tomorrow", labels=["Event"])

# Recall via vector similarity
db.recall(vector=[0.1, 0.5, 0.9], limit=5)

# Vector similarity search
db.similar([0.1, 0.5, 0.9, 0.3], threshold=0.8, limit=10)

# Statistics
db.stats()

API

Method	Description
DriftDB(url, token)	Connect to DriftDB
db.query(driftql)	Execute DriftQL query
db.create_node(labels, properties)	Create a node
db.get_node(id)	Get node by ID
db.list_nodes()	List all nodes
db.delete_node(id)	Soft-delete node
db.find(label, where_clause, returns)	Find by label
db.create_edge(src, tgt, type, props)	Create edge
db.remember(text, labels, vector)	Store AI memory
db.recall(text, vector, limit)	Recall memories
db.similar(vector, threshold, limit)	Vector similarity
db.backup(directory, password)	Backup database
db.health() / db.is_healthy()	Health check
db.stats()	Database stats

Requirements

• Python 3.8+
• A running DriftDB server (driftdb --serve --rest)

License

SSPL-1.0 (same as MongoDB)

GitHub · Documentation