This is basically a fun and learning project.
- In this project, I have implemented a resilient, natural language aware Slack Bot as a micro-service using Node.js, Wit.ai (for natural language processing), Real Time Messaging (RTM) API of Slack, Google’s Geocode and Time-Zone API and Open Weather Map API.
- This Slack Bot provides features of giving date and time of user given location and weather condition of the same.
- Time and Weather are two different microservices (iris-time and iris-weather, here) which are independent of each other as well as of the main application (iris, here).
- This concept of microservices provides resilience to the built system and helps in tackling problems of service failures, e.g. if iris-time is down for some reason, the main application will remain intact and running.
- Clone the repo using git clone command as
git clone https://github.com/amitmangotra/iris-knows-time-and-weather.git - Make sure you have Slack, Google, Wit.ai and Openweathermap accounts.
- You need to have api keys of Google Geocode and TimeZone, Wit.ai, RTM(Real Time Messaging) service of Slack.
- Put the api keys to their respective places in code.
- Create a user-defined bot in Slack and name it
irisand connect it to your slack workspace. - In Wit.ai, configure time and weather intents for training
iris. - Open the folder where you cloned the repo and open iris, iris-time and iris-weather in three separate windows of your favorite IDE.
- Make sure you have NodeJS and npm installed globally.
- Run
npm installon each window in the respective integrated terminals. - In integrated terminal of iris folder, run
npm start - Similarly, for iris-time and iris-weather folders, run
node bin/run.jsin their respective integrated terminals. - Open #general channel of your slack workspace and test iris-time and iris-weather services using queries like, "How's the weather in Boston, Iris?" or "What's the time in Sydney, Iris?".Make sure each query must contain the word "iris".