Problem/Solution statement
*note: due to limited demo time only a few features were demoed. For the full trained dataset, check the json file embed with this submission. There are ~95 intents coded with an average of 8 examples per intent due to limited time.
AXA
Problem:
Innovative solutions for protecting what matters such as pandemic aid, biodiversity, catastrophe detection/recovery, fighting mental illness,...
Solutions:
With the pandemic, the whole world has been turned upside down. To help us facilitate the return to normal, I made a digital assistant that is easy to use, intuitive and can be deployed across many platforms. LauzBot can do tasks (send reminder emails,etc), grab info very quickly (check bank account balance, bank routing number, current COVID cases,etc) or be a friend (tell jokes, suggest places to eat or things to do,etc). LauzBot can perform a symptom check, and reccomend the best course of action based on your answers to the questionaire. with the pandemic, depression rates has skyrocketed as we get further apart from less contact. LauzBot was made to simulate a friend. LauzBot also has sentiment analysis and can detect when you're sad. Lauz bot can also listen to you vent your frustrations. LauzBot can also tell jokes, suggest you things to do or places to eat based on your preferences (pretrained data that will have to be customized/setup user to user).
OpenWebTechnology
Problem:
The digital revolution has massively changed our private everyday life. Why don't we also benefit from this in our daily work?
Solution:
Today, employees lose a lot of time with administrative tasks and searching for information in different system. Digital assistants provide easy access to information and allow efficient completion of administrative tasks.They also promote exchange between employees. The problem presented a choice between a chatbot or employee app. I made a chatbot that can be deployed to employee app (which was demo deployed to Slack). It is a digital assistant that is easy to use, intuitive and can be deployed across many platforms. LauzBot can do tasks (send reminder emails,etc), grab info very quickly (bank routing number, current COVID cases,etc) or be a friend (tell jokes, suggest places to eat or things to do,etc).
Try it out
Test out the bot by downloading the html file and running it, or clicking this link and hit run: https://www.w3schools.com/code/tryit.asp?filename=GL6ZXSK4M9KA
Tracks targetting
AXA and OpenWebTechnology
How I built it
Java on Android studio for android app deployment, Node.js 10 for functionalities, Watson API for machine learning, speech to text, text to speech, tone and sentimental analysis
Challenges I ran into
Making a dataset from scratch and training the bot on that dataset, working with Watson, deploying it as an Android app, Slack bot and embeded web code all in one shot.
Accomplishments that I'm proud of
Doing all this in a short amount of time. First time working with AI and ML
Log in or sign up for Devpost to join the conversation.