Chatbot Training Data

API which manages user logs and adds new queries and their answers to chatbot-api

Table Of Contents

• Overview
• Development
• Deployment

Overview

HTTP RESTful API.

Requests pass data via JSON encoded bodies except for in GET requests where data will be passed via URL and query parameters.

Responses will always return JSON String.

Development

The Chatbot-training-data server can be run locally. Visit DESIGN.md to see all endpoints and corresponding responses.

Follow the steps in the Database, Configuration, and Run sections.

Database

Start a local MongoDB server by running:

make db

Configuration

Configuration is passed via environment variables.

• API_KEY : API key assigned to the bot
• WORKSPACE_ID :Unique id given to the created skill
• APP_DB_HOST : Database host
• APP_DB_NAME : Database name
• APP_DB_USER : User for this database
• APP_DB_PASSWORD : Database user password

Deployment

Deployment Configuration

Create a copy of deploy/values.secrets.example.yaml named deploy/values.secrets.ENV.yaml for whichever deployment environment you wish to configure.

Edit this file with your own values.

Never commit this file.

Deploy

Initialize submodules:

git submodule update --init --recursive

Deploy production by running:

make deploy-prod

If this is the first time production has been deployed run:

make rollout-prod

Access

In order to access the ChatBot Training Data, you need to be logged in to the cluster from your terminal.

To access the API on local machine, Start the proxy to the cluster, use command:

make proxy

In order to check if the proxy is running and cluster can be accessed, use:

make get-health

To access API, call the URL on the browser or Postman :

http://localhost:8001/api/v1/namespaces/kscout/services/<ENV>-chat-bot-training-data:http/proxy/<ENDPOINT>

Replace <ENV> with an environment.
Replace <ENDPOINT> with the endpoint you want to access.