Hive Forensics AI - Vektaris API Client

🚀 Introduction

Welcome to the Hive Forensics AI API client for interacting with Vektaris, our cutting-edge vector database built on the ICP blockchain. This npm package provides a simple and efficient way for developers to integrate our API into their applications.

Vektaris enables seamless storage, retrieval, and querying of vector embeddings, optimized for AI-driven applications. Our current beta phase is open for testers—sign up now to gain early access!

🌐 Project Links

• 🌎 Hive Forensics AI: https://www.hiveforensics.com
• 🔬 Vektaris (Vector DB Engine): https://www.vektaris.com
• 🛠 Developed by: Samuel Paniagua

---

📦 Installation

To install the Vektaris API client, run the following command:

npm install vektaris-client

Alternatively, if cloning from our repository:

git clone https://github.com/hiveforensics/vektaris-client.git
cd vektaris-client
npm install
npm run build

---

⚙️ Configuration

Before using the package, create a .env file to store your API credentials:

VEKTARIS_BASE_URL=https://api.vektaris.com
VEKTARIS_USER_KEY=your-api-key-here
VEKTARIS_DB_NAME=your-database-name
VEKTARIS_ACCOUNT=your-email@example.com

Ensure you replace the placeholders with your actual credentials after signing up.

---

🔥 Features

✅ Create Databases (Namespaces)
✅ Store Vector Embeddings (one at a time due to ICP constraints)
✅ Run Semantic Queries
✅ Retrieve Stored Embeddings (for external LLMs or AI models)
✅ Perform Hybrid Search (text-based & semantic similarity combined)

---

📖 Usage Guide

1️⃣ Import & Initialize Client

import VektarisClient from 'vektaris-client';

const client = new VektarisClient(); // Uses .env values by default

2️⃣ Create a New Database

await client.createDatabase({
  dbName: 'MyVectorDB',
  account: 'user@example.com',
});

3️⃣ Store an Embedding

await client.storeEmbedding({
  dbName: 'MyVectorDB',
  account: 'user@example.com',
  title: 'AI-driven cybersecurity solutions',
  content: 'This document discusses AI models for cyber threat detection...'
});

4️⃣ Query Embeddings

const results = await client.queryEmbeddings({
  dbName: 'MyVectorDB',
  account: 'user@example.com',
  query: 'cybersecurity threats'
});
console.log(results);

5️⃣ Retrieve All Embeddings (No Query)

const embeddings = await client.getEmbeddings({
  dbName: 'MyVectorDB',
  account: 'user@example.com',
  start: 0,
  limit: 100
});
console.log(embeddings);

6️⃣ Hybrid Search

const hybridResults = await client.hybridSearch({
  dbName: 'MyVectorDB',
  account: 'user@example.com',
  query: 'AI in security',
  limit: 10
});
console.log(hybridResults);

---

🚧 Beta Testing & Feedback

We are currently in our beta phase, and we value your feedback! If you encounter any issues or have feature suggestions, please visit our website and contact us.

🔗 Sign up for early access: https://www.vektaris.com

---

💡 Future Enhancements

• 🔄 Batch Upload Support (Store multiple embeddings in sequence)
• 🛡️ Security Enhancements (Encryption for stored embeddings)
• 📊 Advanced Analytics Dashboard (Monitor vector similarity trends)

---

📜 License

This project is licensed under the MIT License.

---

👨‍💻 Developed By

👤 Samuel Paniagua
🌐 Portfolio
🐦 Twitter

---

⭐ If you find this package useful, consider giving it a star on GitHub!