Repolode
Submit
SEPIA-Framework

SEPIA-Framework/sepia-assist-server

Core server of the SEPIA voice assistant framework, handling NLU, conversation flow, smart-service integration and user accounts via a REST API.

View on GitHubWebsite
Stars
101
Forks
16
Open issues
0
Last commit
23 days ago

What is it?

SEPIA Assist-Server is the backend "brain" of the SEPIA Framework, a self-hosted personal voice assistant. It bundles several modules behind a RESTful Assist-API: user-account management, database integration (for example Elasticsearch), natural-language understanding and named-entity recognition, a conversation flow module, an answer module, and smart-services that connect local services such as to-do lists or cloud APIs like weather to the assistant.

The NLU works out of the box for German and English, and its modular chain can call external APIs and Python scripts. Text-to-speech is integrated with eSpeak, MaryTTS and picoTTS, and can run either on this server or inside the SEPIA client. Remote actions let IoT devices or wake-word tools send data to the server.

SEPIA clients access the API directly for authentication and data exchange, or connect to the SEPIA chat server to exchange messages across channels with multiple users or devices. The server is designed to run the same way on a Raspberry Pi for a small private network or on multiple servers for a larger deployment.

Key features

  • 01Natural-language understanding and named-entity recognition, working out of the box for German and English
  • 02Conversation flow and answer modules exposed through a RESTful Assist-API
  • 03Smart-service integration for local services and cloud APIs such as weather
  • 04Built-in user-account management and database integration such as Elasticsearch
  • 05Text-to-speech through eSpeak, MaryTTS or picoTTS

Use cases

  • Running a private, self-hosted voice assistant on a Raspberry Pi for a small household or group
  • Connecting IoT devices or wake-word tools to a central assistant backend
  • Building custom NLU pipelines that call external APIs or Python scripts
  • Hosting a multi-user assistant backend for a larger company network

Things to know

  • The README gives no install or run commands, so setup steps are not listed here.
  • Requires self-hosting and a Java-based deployment for the server.

In collections

Self-hosted appsAI chatbots & chat clients

Topics

#voice-assistant#chatbot#nlu#self-hosted#java#raspberry-pi#smart-home#rest-api#ai#assist-api

FAQ

What is sepia-assist-server?
Core server of the SEPIA voice assistant framework, handling NLU, conversation flow, smart-service integration and user accounts via a REST API. SEPIA Assist-Server is the backend "brain" of the SEPIA Framework, a self-hosted personal voice assistant. It bundles several modules behind a RESTful Assist-API: user-account management, database integration (for example Elasticsearch), natural-language understanding and named-entity recognition, a conversation flow module, an answer module, and smart-services that connect local services such as to-do lists or cloud APIs like weather to the assistant.
Is sepia-assist-server free to use? What license does it have?
sepia-assist-server is open source under the MIT license. Check the license terms before using it in commercial products.
Is sepia-assist-server still maintained?
Actively maintained. The last commit to sepia-assist-server was on Sep 29, 2026.
What can I use sepia-assist-server for?
Running a private, self-hosted voice assistant on a Raspberry Pi for a small household or group Connecting IoT devices or wake-word tools to a central assistant backend Building custom NLU pipelines that call external APIs or Python scripts Hosting a multi-user assistant backend for a larger company network
Where can I get sepia-assist-server?
The source code and installation instructions are on GitHub: https://github.com/SEPIA-Framework/sepia-assist-server

Summary generated by AI from the project's README. GitHub stats refresh daily.

Related projects