Track: VoiceTech |
HAL: A Conversational AI Agent for Smart-Home Automation (Live Demo) |
One of the biggest problems with smart-home devices currently, is the need to issue precise commands, or adapt your voice to be more mechanical, or use smaller voice commands so that an IoT device can understand you. One may also be limited to pressing a button or gesturing in order to get a response from an IoT device. HAL is a voice enabled conversational AI agent, that brings natural language understanding to industrial-scale IoT. It leverages machine learning, intent recognition and Natural Language Understanding (NLU) to think, act and respond to user queries, thereby providing a human-like interface to IoT devices. HAL is omni-channel and supports multiple channels of communication, including voice, webchat, Facebook Messenger, etc. It is context-aware and considers the current state of a user conversation, along with the incoming user message, to make an intelligent decision, based on a neural network model. HAL has a microservices architecture, making it highly scalable, resilient and easy to maintain. In this talk, we intend to show a live demo of HAL. We will talk to HAL in real-time, and show how it seamlessly integrates with smart-home IoT devices, while providing a natural conversational experience. |
|
Presentation Video |
Presentation Notes |
Sharma-HAL-A-Conversational-AI-Agent.pdf |