Say Hello to the new wave of software bots... by Irikidzai Muchaneta

No, really, say "hello" to them. The age of voice enabled interfaces is already upon us with the rise of driverless cars, ubiquitous technology and the Internet of Things(IoT), computers no longer need to speak in code but they have learnt to speak the human language. Google home is a voice-activated speaker powered by the google assistant similar to other assistants namely Siri, Alexa and Cortana.

In this talk, I will be looking at developing and deploying Actions (Conversations) on the Google home assistant. This will be done by creating an application (web hook) triggered by an action created in API.AI using a python template.

* Audience is anyone interested in creating bots using API.AI
* Anyone interested in adding customized actions to the google assistant
* Anyone interested in learning about voice enabled actions

Specific aspects I will cover include:
- Reasons, voice enabled interfacing is an interesting topic to follow:
- Simplified overview of the steps to create custom "conversation actions"
- Going into a bit of detail where the challenges arise
- Examples of cool projects out there currently

Transcript

1. TITLE: Say hello to the next wave of software bots…. IRIKIDZAI MUCHANETA (Ikky) PyConZA2017 – October 2017 1
5. Iron Man speaking to Jarvis
6. The Goal: “To get from point A (Today) to point B (iron man)” Focus of Talk: How to Create applications/Actions for the Google Assistant which in turn integrate with our python programs via an API.
7. Python Program / Code Web Interface / API NLU / API.AI / Google Actions Google Home The Structure:
8. Reasons for Developing Conversational Apps: - Internet of Things - Blue Ocean Strategies - New medium to reach customers The Why
9. 1. Overview 2. Creating Actions / Agents in API.AI 3. Creating Intents and Entities (Main concepts) 4. Testing our conversations in the API.AI simulator 5. Add python integrations to our WebHook The Steps
10. How Do Google Actions Work 1. Overview
11. Extra Layer 1. Overview
12. How it fits together 1. Overview
13. 1. Actions Console
14. 2. API.AI Console
15. 2. API.AI Console Systematic order of Execution points: Agents Intents Entities Integrations Fulfilment
16. 2. API.AI Console
17. Requests / Queries 3. Intents
18. Keywords / Search Terms 4. Entities
19. 3. Intents
22. Summarize Think of an example of data that you would like to ask the system. Order it into the following: - Request / Intent - Key Word searches / Entities - Create Actions - Test the Actions - Get results and deploy PyConZA2017 – October 2017 22
23. Questions