Skip to content

Software and Skills for Personal Assistants using SUSI AI / Artificial Intelligence for Chat Bots http://api.susi.ai

License

Notifications You must be signed in to change notification settings

derekfrost12140/susi_server

 
 

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Susi

Join the chat at https://gitter.im/fossasia/susi_server Build Status Percentage of issues still open Average time to resolve an issue Twitter Twitter Follow

Susi AI is an intelligent Open Source personal assistant. It is capable of chat and voice interaction by using APIS to perform actions such as music playback, making to-do lists, setting alarms, streaming podcasts, playing audiobooks, and providing weather, traffic, and other real time information. Additional functionalities can be added as console services using external APIs. Susi AI is able to answer questions and depending on the context will ask for additional information in order to perform the desired outcome. The core of the assistant is the Susi AI server that holds the "intelligence" and "personality" of Susi AI. The Android and web applications make use of the APIs to access information from a hosted server.

An automatic deployment from the development branch at GitHub is available for tests here https://susi-server.herokuapp.com

Communication

Please join our mailing list to discuss questions regarding the project: https://groups.google.com/forum/#!forum/opntec-dev/

Our chat channel is on gitter here: https://gitter.im/fossasia/susi_server

How do I install Susi: Download, Build, Run

Note

  • You must be logged in to Docker Cloud for the button to work correctly. If you are not logged in, you'll see a 404 error instead.

Deploy Deploy on Scalingo Deploy to Bluemix Deploy to Docker Cloud

At this time, Susi AI is not provided in compiled form, you easily build it yourself. It's not difficult and done in one minute! The source code is hosted at https://github.com/fossasia/susi_server, you can download it and run Susi AI with:

> git clone https://github.com/fossasia/susi_server.git
> cd susi_server
> git submodule update --recursive --remote
> git submodule update --init --recursive
> ./gradlew build
> bin/start.sh

For Windows Users (who are using GitBash/Cygwin or any terminal):

> git clone https://github.com/fossasia/susi_server.git
> cd susi_server
> git checkout master
> ant jar
> java -jar dist/susiserver.jar
> git checkout development
> ant jar
> java -jar dist/susiserver.jar
To stop:
> Press Ctrl+C

After all server processes are running, Susi AI tries to open a browser page itself. If that does not happen, just open http://localhost:4000; if you made the installation on a headless or remote server, then replace 'localhost' with your server name.

To stop Susi AI, run: (this will block until the server has actually terminated)

> bin/stop.sh

A self-upgrading process is available which must be triggered by a shell command. Just run:

> bin/upgrade.sh

Where can I download ready-built releases of Susi AI?

No-where, you must clone the git repository of Susi AI and built it yourself. That's easy, just do

  • git clone https://github.com/fossasia/susi_server.git
  • cd susi
  • then see below ("How do I run Susi AI")

How do I install Susi AI with Docker on Google Cloud?

To install Susi AI with Docker on Google Cloud please refer to the Susi Docker installation readme.

How do I install Susi AI with Docker on AWS?

To install Susi AI with Docker on AWS please refer to the Susi Docker installation readme.

How do I install Susi AI with Docker on Bluemix ?

To install Susi AI with Docker on Bluemix please refer to the Susi Docker installation readme.

How do I install Susi AI with Docker on Digital Ocean ?

To install Susi AI with Docker on Digital Ocean please refer to the Susi Docker installation readme.

How do I deploy Susi AI with Heroku?

You can easily deploy to Heroku by clicking the Deploy to Heroku button above. To install Susi AI using Heroku Toolbelt, please refer to the Susi Heroku installation readme.

How do I deploy Susi AI with cloud9?

To install Susi AI with cloud9 please refer to the Susi cloud9 installation readme.

How do I setup Susi AI on Eclipse?

To install Susi AI on Eclipes, please refer to the Susi Eclipse readme.

How do I run Susi AI?

  • build Susi (you need to do this only once, see above)
  • run bin/start.sh
  • open http://localhost:4000 in your browser
  • to shut down Susi, run bin/stop.sh

How do I configure Susi AI?

The basis configuration file is in conf/config.properties. To customize these settings place a file customized_config.properties to the path data/settings/

How to compile using Gradle?

  • To install Gradle on Ubuntu:

    $ sudo add-apt-repository ppa:cwchien/gradle
    $ sudo apt-get update
    $ sudo apt-get install gradle
    
  • To install Gradle on Mac OS X with homebrew

    brew install gradle
    
  • To compile, first, create dir necessary for Gradle

    ./gradle_init.sh
    

    Compile the source to classes and a jar file

    gradle assemble
    

    Compiled file can be found in build dir Last, clean up so that we can still build the project using Ant

    ./gradle_clean.sh
    

How do I develop Skills (AI Conversation Rules) for Susi AI?

The Susi AI skill language is described in the Skill Development Tutorial.

Why should I use Susi AI?

If you like to create your own AI, then you may consider Susi AI.

Where can I get the latest news about Susi AI?

Hey, this is the tool for that! Just put http://loklak.org/api/search.rss?q=%23susi into your rss reader. Oh wait.. you will get a lot of information about tasty Cambodian food with that as well. Alternatively you may also read the authors timeline using http://loklak.org/api/search.rss?q=0rb1t3r or just follow @0rb1t3r (that's a zero after the at sign)

Where can I find documentation?

The Doumentation can be found at https://github.com/fossasia/susi_server/tree/development/docs/api/api.md

Where do I find the javadocs?

You can build them via 'ant javadoc'

Where can I report bugs and make feature requests?

This project is considered a community work. The development crew consist of you too. I am very thankful for pull request. So if you discovered that something can be enhanced, please do it yourself and make a pull request. If you find a bug, please try to fix it. If you report a bug to me I will possibly consider it but at the very end of a giant, always growing heap of work. The best chance for you to get things done is to try it yourself. Our issue tracker is here.

What is the software license?

LGPL 2.1

Development Workflow

Fixing issues

Step 1: Pick an issue to fix

After selecting the issue

1.Comment on the issue saying you are working on the issue.

2.We expect you to discuss the approach either by commenting or in the gitter.

3.Updates or progress on the issue would be nice.

Step 2: Branch policy

Start off from your development branch and make sure it is up-to-date with the latest version of the committer repo's development branch. Make sure you are working in development branch only. git pull upstream development

If you have not added upstream follow the steps given here.

Step 3: Coding Policy

  • Please help us follow the best practice to make it easy for the reviewer as well as the contributor. We want to focus on the code quality more than on managing pull request ethics.
  • Single commit per pull request
  • For writing commit messages please adhere to the Commit style guidelines.
  • Follow uniform design practices. The design language must be consistent throughout the app.
  • The pull request will not get merged until and unless the commits are squashed. In case there are multiple commits on the PR, the commit author needs to squash them and not the maintainers cherrypicking and merging squashes.
  • If you don't know what does squashing of commits is read from here.
  • If the PR is related to any front end change, please attach relevant screenshots in the pull request description

Step 4: Submitting a PR

Once a PR is opened, try and complete it within 2 weeks, or at least stay actively working on it. Inactivity for a long period may necessitate a closure of the PR. As mentioned earlier updates would be nice.

Step 5: Code Review

Your code will be reviewed, in this sequence, by:

  • Travis CI: by building and running tests. If there are failed tests, the build will be marked as a failure. You can consult the CI log to find which tests. Ensure that all tests pass before triggering another build.
  • The CI log will also contain the command that will enable running the failed tests locally.
  • Reviewer: A core team member will be assigned to the PR as its reviewer, who will approve your PR or he will suggest changes.

Have fun! @0rb1t3r

About

Software and Skills for Personal Assistants using SUSI AI / Artificial Intelligence for Chat Bots http://api.susi.ai

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages

  • Java 82.1%
  • HTML 11.3%
  • JavaScript 3.8%
  • CSS 1.6%
  • Shell 1.2%
  • Batchfile 0.0%