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.
Deployments
Development: An automatic deployment from the development branch at GitHub is available for tests at https://susi-server.herokuapp.com
Master: The master branch is planned to be deployed on https://api.susi.ai. Currently, the deployment is taking place each hour at xx.45 using the development branch. We are planning to switch to the Master branch for production soon.
Deployment branches
There are two branches targetting deployment, "dev-dist" and "stable-dist". The former is intended to be used in conjunction with the development versions of "susi_linux", while the later with the stable (from the "master" branch). These two branches are currently updated manually from "susi_server_binary_latest.tar.gz" (see below), but future integration will update "dev-dist" regularly.
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
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.
At this time, SUSI.AI is not provided in the 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 (Before installation you must have "Java Development Kit" latest version at http://openjdk.java.net/install/ & "Gradle" latest version at https://gradle.org/install/):
Note
- For Armv6 processors (e.g. Raspberry Pi Zero / Zero W/ Zero WH/ 1A / 1B), please make sure that your system is using Java 8 (Oracle or OpenJDK) as there are some compatibility issues for Armv6 processors.
- You may use the following command to install OpenJDK's Java 8 JRE and JDK:
> sudo apt install openjdk-8-jdk-headless
> git clone https://github.com/fossasia/susi_server.git > cd susi_server > ./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
The latest binary built can be downloaded from http://download.susi.ai/susi_server/susi_server_binary_latest.tar.gz
To run susi, do: tar xfz susi_server_binary_latest.tar.gz cd susi_server_binary_latest java -server -Xmx200m -jar build/libs/susi_server-all.jar
To install SUSI.AI with Docker on Google Cloud please refer to the Susi Docker installation readme.
To install SUSI.AI with Docker on AWS please refer to the Susi Docker installation readme.
To install SUSI.AI with Docker on Bluemix please refer to the Susi Docker installation readme.
To install SUSI.AI with Docker on Azure please refer to the Susi Docker installation readme.
To install SUSI.AI with Docker on Digital Ocean please refer to the Susi Docker installation readme.
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.
To install SUSI.AI with cloud9 please refer to the Susi cloud9 installation readme.
To install SUSI.AI on Eclipse, please refer to the Susi Eclipse readme.
- build Susi (you need to do this only once, see above)
- run
bin/start.sh - open
http://localhost:4000in your browser - to shut down Susi, run
bin/stop.sh
The basis configuration file is in conf/config.properties. To customize these settings place a file customized_config.properties to the path data/settings/
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
The compiled file can be found in build dir Last, clean up so that we can still build the project using Ant
./gradle_clean.sh
The SUSI.AI skill language is described in the Skill Development Tutorial.
If you simply want to add your skill to the SUSI.AI online service, please go to https://skills.susi.ai and add your skill.
For your own deployments: The Susi skill data is the storage place for the Susi skills. To make Susi server utilize these skills, clone Susi skill data alongside Susi server.
git clone https://github.com/fossasia/susi_skill_data.git
If you want to create private skills in your local server, you should create a local git repository susi_private_skill_data alongside Susi server. Then you must create a local git host:
> cd <above susi home> > mkdir susi_private_skill_data_host > cd susi_private_skill_data_host > git init —bare > cd ../susi_private_skill_data > git remote add origin <path to susi_private_skill_data_host> > git push --set-upstream origin master
If you like to create your own AI, then you may consider SUSI.AI.
The Documentation can be found here.
You can build them via 'ant javadoc'
This project is considered a community work. The development team consists of you too. We are very thankful for the 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 us, We 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.
After selecting the issue
- Comment on the issue saying you are working on the issue.
- We expect you to discuss the approach either by commenting or on Gitter Chat.
- Updates or progress on the issue would be nice.
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.
- 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 cherry-picking 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
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.
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.