README.md

ZeroKDB

ZeroKDB is a powerful Python library designed to facilitate seamless interaction with the ZeroKDB API and to create your own decentralized database. It provides an easy-to-use interface for executing queries and generating proofs.

Features

• Seamless Database Interaction: Connect and interact with ZeroKDB using simple API calls.
• AI Model Integration: Convert text to embeddings and execute similarity queries.
• Proof Generation: Generate cryptographic proofs to validate your queries.
• Support for IPFS Storage: Utilize IPFS for decentralized storage solutions.

Installation

ZeroKDB can be installed by adding it as a dependency in your project.

pip install git+ssh://[email protected]/agencyenterprise/zerokdb.git

Installation

Below is a basic example to get you started with ZeroKDB. This example demonstrates how to instanciate a Database API instance and use it to execute queries with proof validations.

• choose between ipfse storage or local storage for the DB
• need to have a pinata key for interacting with ipfs
• need to reference the api where ZerokDB API is running and keeping track of sequences

import os
from zerokdb.api import DatabaseAPI
import json

async def generate_proof(
    proof_request_id: str, ai_model_name: str, ai_model_inputs: str, pina_api_key: str
):
    print("Generating proof for request " + proof_request_id)
    result = None

    if ai_model_name != "zerokdb":
        return None, None
    db_api = DatabaseAPI(
        storage_type="ipfs",
        pinata_api_key=pina_api_key,
        api_host=os.getenv("API_HOST") or "http://localhost:8001"
    )
    ai_model_inputs_dict = json.loads(ai_model_inputs)

    if ai_model_inputs_dict["type"] == "TEXT":
        text = ai_model_inputs_dict["value"]["text"]
        table_name = ai_model_inputs_dict["value"]["table_name"]
        vector = db_api.convert_text_to_embedding(text)
        sql_query = f"SELECT * FROM {table_name} LIMIT 5 COSINE SIMILARITY embedding WITH {vector}"
    elif ai_model_inputs_dict["type"] == "SQL":
        sql_query = ai_model_inputs_dict["value"]
    else:
        raise ValueError("Unsupported AI model input type")

    result, circuit, proof = db_api.execute_query(sql_query, proof=True)

    print("Proof generated for request: ", proof_request_id)
    return circuit, proof, result