SensibleDB

Embedded Graph-Vector Database for AI Applications

SensibleDB is a unified database that makes it easy to build all the components needed for an AI application in a single platform. You no longer need a separate application DB, vector DB, graph DB, or application layers managing multiple storage locations.

Built in Rust with LMDB as its storage engine — like SQLite for knowledge graphs.

---

Choose Your Path

I want to...	Start here
Organize my personal documents and search through them using AI	SensibleDB Explorer — Desktop app for non-technical users
Build AI applications with a knowledge base	CLI Quick Start — For developers
Store memory for AI agents	Embedded Mode — For Rust developers

---

SensibleDB Explorer — Your Personal Knowledge App

For non-technical users who want to organize and search their personal documents.

Why SensibleDB Explorer?

• 100% Local-First — Your documents never leave your computer. No cloud, nothing sent externally.
• Works Offline — Disconnect your internet and keep using the app.
• Chat with Your Documents — Ask questions in plain English, get answers from your files.

Get Started

1. Download SensibleDB Explorer for macOS
2. Launch the app
3. Click "Add Folder" → select your documents folder
4. Start chatting!

---

CLI Quick Start — For Developers

1. Install the CLI (build from source)

git clone https://github.com/Sensible-Analytics/SensibleDB.git
cd SensibleDB
cargo install --path sensibledb-cli

Note: The install script is temporarily unavailable. Use source build above.

2. Initialize a Project

mkdir my-project && cd my-project
nexus init

3. Write Your Schema & Queries

Open the generated .hx file and define your schema:

N::User {
    INDEX name: String,
    age: U32
}

E::FRIENDS {
    from: User,
    to: User
}

QUERY getUser(user_name: String) =>
    user <- N<User>({name: user_name})
    RETURN user

QUERY getFriends(user_id: U64) =>
    user <- N<User>({id: user_id})
    friends <- TRAVERSE user -[FRIENDS]-> friend
    RETURN friend

4. Deploy

nexus push dev

5. Use in Your Application

TypeScript (use HTTP directly):

// SDK coming soon - use HTTP for now
const response = await fetch("http://localhost:6969/getUser", {
  method: "POST",
  headers: { "Content-Type": "application/json" },
  body: JSON.stringify({ user_name: "John" })
});
const user = await response.json();

Python (use requests directly):

import requests

response = requests.post(
    "http://localhost:6969/getUser",
    json={"user_name": "John"}
)
user = response.json()
print(user)

Note: TypeScript and Python SDKs are in development. Track progress on GitHub.

---

Embedded Mode — For Rust Developers

For full control, embed SensibleDB directly in your application:

[dependencies]
sensibledb-db = { version = "1.3", features = ["embedded"] }

use sensibledb_db::embedded::{Database, Node};

let db = Database::open("./my_db")?;
let mut tx = db.write_transaction()?;

tx.put_node(Node {
    id: 1,
    label: "User".into(),
})?;

tx.commit()?;

---

Key Features

• 100% Local-First — Runs entirely on your machine, no external calls
• Built-in Embeddings — Vectorize text directly in queries with Embed()
• RAG Tooling — Vector search, keyword search, and graph traversals out of the box
• Secure by Default — Private by default; access only through compiled SensibleQL queries
• Ultra-Low Latency — Rust + LMDB storage engine for extreme performance
• Type-Safe Queries — SensibleQL is 100% type-safe, catch errors at compile time
• MCP Ready — (Coming Soon) Connect AI agents via MCP server

---

Documentation

Section	Description
Getting Started	Installation and setup
SensibleQL	Query language reference
CLI	Command-line tool guide (called "nexus")
SDKs	TypeScript, Python, and Rust SDKs
Features	Full feature overview
Programming Interfaces	Rust embedded API in 5 minutes

---