Wassim Chegham PRO
Senior Developer Advocate @Microsoft ★ Angular contributor ★ Bazel contributor ★ GDE @Google ★ creator of @itsjustangular / hueaction.dev / ngx.tools / xlayers.dev / angular.run / thundr.dev
Before We Begin...
* remember the details of the conversation
core brand attributes
voice, style and dialog
Canonical “Happy path”
"Ok Google, talk to bot"
"Sure! Here is bot"
Invoke an action
"Hi! Welcome to bot. i can help you do..."
"I want to hear more about Chatbots"
Speech to Text, NLP, Knowledge Graph, ML Ranking, User Profile Text to Speech
Speech to Text
Text to speech
A "serverless platform for building event-based microservices".
Direct HTTP calls
Google Analytics for Firebase*
Debug in Chrome
# With gcloud $ gcloud beta functions deploy agent \ --trigger-http \ --stage-bucket gs://my-agent/ # With Firebase $ firebase deploy --only functions # Deploy Locally $ functions deploy agent --trigger-http
→Merchandising and information
→Automatic and manual policy checks
→Turn around in about 1 week
Actions on Google
Google Home, TV, CARS… and more to come
Slack, Facebook Messenger, Twitter, Twilio, Skype, Tropo, Telegram, Kik, LINE, Cisco Spark, Alexa, Cortana...
By Wassim Chegham
Ever wanted to build a chatbot? What if you could use a voice assistant to manage your agenda for the conference? Know if there are any presentations on Machine Learning, or Docker, or your favorite programming language? Find out more about a speaker...etc. In this session, we will look at how to build our own voice assistant, using Google Cloud's natural speech and voice recognition APIs with DialogFlow services, and Google Cloud Functions to implement necessary business logic.