Skip to content

README_EN.md

Latest commit

 

History

History
108 lines (79 loc) · 4.16 KB

README_EN.md

File metadata and controls

108 lines (79 loc) · 4.16 KB

Augo Logo

Augo

Self-Hosted AI Financial Assistant

License: AGPL v3 Python 3.13 Flutter PRs Welcome

简体中文 | English

Augo is an open-source AI financial assistant that supports self-hosting. It helps you manage personal or family finances while keeping data and AI memory on your own infrastructure. It can be flexibly connected to various cloud or local LLMs.

🛠️ Features

Agent-Based Financial Processing

Powered by LangGraph, Augo goes beyond simple transaction logging. It understands your intent through natural conversation, supporting multi-step financial queries, self-correction, and long-term context retention.

Model-Agnostic Connectivity

Augo doesn't lock you into a single AI provider:

  • Cloud Models: Connect to OpenAI, DeepSeek, and other mainstream providers.
  • Local Models: Use Ollama or other frameworks to run models on your own hardware for maximum privacy.

Voice & Privacy

Integrates with the asr_server project (customized from open-source ASR solutions):

  • Self-Hosted ASR: Deploy your own voice recognition service.
  • Switchable Engines: Easily switch between system-native and local ASR services to keep voice data within your network.

Skill-Based Extensions

Uses a plugin mechanism inspired by Anthropic Skills. You can add specialized Skill scripts to give the assistant advanced capabilities like spending analysis or budget planning.

User Interaction

Implements the Google A2UI protocol, where UI components (charts, detail cards) are pushed dynamically by the AI based on the conversation context. The frontend is built with the forui UI library.

🏠 Deployment Scenarios

  • NAS Self-Hosting: Optimized for Synology, QNAP, and other common NAS systems using one-click Docker deployment.
  • Family Sharing: Supports multiple members with centralized data storage under your full control.
  • Developer Friendly: Includes a full Makefile toolchain for local development.

⚙️ Tech Stack

  • Backend: Python 3.13 + FastAPI, with uv for dependency management.
  • Orchestration: LangGraph + Mem0.
  • Frontend: Flutter + forui.
  • Storage: PostgreSQL (pgvector) for financial records and vector data.

🚀 Quick Start

Prerequisites

  • Docker & Docker Compose (Recommended)
  • AI Model API Key or Local LLM endpoint

🐳 Docker Deployment

  1. Clone the repo: 
    git clone https://github.com/kylesean/augo.git && cd augo
  2. Configure Environment: 
    cp server/.env.example server/.env
# Update server/.env with your Keys or local endpoint
  3. Start: 
    make docker-up
    Scan the terminal QR code with your mobile app to connect.

🗺️ Project Structure

  • /client: Flutter client source code.
  • /server: FastAPI backend source code.
  • /docker-compose.yml: Docker orchestration configuration.

📄 License

Licensed under AGPLv3.

Email: jkxsai@gmail.com | WeChat: Ky1eSean