Chatbot with large DB integration - bots

Question

We want to build a Bot, which will ask questions to user and then
will record their response. This Q&A will be intelligently done based upon responses from user.
We have few initial questions at this moment;
There has been lots of API's available for bot creation like api.ai, wit.ai, botkit, IBM Watson, Microsoft Bot framework. Which one is best from development point of view in Java or Node.Js or Python. Also, from the point of having control on the data and complete flow.
We want to provide the questions from some DB[RDBMS or NOSQL] to our Bot as it will be large in future, which api would be best for the same.
We want to store the user response to the database with userid.
Based upon user response stored in DB, to the questions asked by bot we want to perform the analytics on it.
Can you please suggest, if this can be done using any one of the Bot API and which one should be preferred.
Thank you,
Amit

Answer 1

In your 4 conditions, all can be done using IBM Watson.
1:
Use Conversation Service for create a Chatbot, and you can save all user input with context variables.
IBM Watson provided some examples in Python, Node JS and Java SDK, just click in some Programming language for check the example and all codes.
2:
This example use Cloudant DB (nosql) from Conversation Simple Node.js link, but you can use other.
function log(input, output) {
if ( logs ) {
// If the logs db is set, then we want to record all input and responses
var id = uuid.v4();
logs.insert( {'_id': id, 'request': input, 'response': output, 'time': new Date()} );
}
}
if ( cloudantUrl ) {
// If logging has been enabled (as signalled by the presence of the cloudantUrl) then the
// app developer must also specify a LOG_USER and LOG_PASS env vars.
if ( !process.env.LOG_USER || !process.env.LOG_PASS ) {
throw new Error( 'LOG_USER OR LOG_PASS not defined, both required to enable logging!' );
}
// add basic auth to the endpoints to retrieve the logs!
var auth = basicAuth( process.env.LOG_USER, process.env.LOG_PASS );
// If the cloudantUrl has been configured then we will want to set up a nano client
var nano = require( 'nano' )( cloudantUrl );
// add a new API which allows us to retrieve the logs (note this is not secure)
nano.db.get( 'car_logs', function(err) {
if ( err ) {
console.error( err );
nano.db.create( 'car_logs', function(errCreate) {
console.error( errCreate );
logs = nano.db.use( 'car_logs' );
} );
} else {
logs = nano.db.use( 'car_logs' );
}
} );
3: All call conversation have some id, you can access this with context variable. Example (With IBM Watson conversation:
context.conversation_id
4: You can use other service from IBM Watson for it, but i recommend: AlchemyAPI or Discovery, depends on what you are really going to do. But take a look at both that I'm sure they poed to help you.