Inspiration
1 in 5 people experience a mental health issue every year 8 people die every day in Australia more than road deaths or cancer 25% of 1M+ people calling support services are not answered
What it does
Amelie is an artificial intelligent chatbot that can help people discover how to improve their mental health; Mental health providers can use Amelie to triage their call volume and identify people at risk. Amelie can also help check in on people who have been discharged but are still at risk.
How we built it
Developed conversational search capability to discover domain specific content. Deployed to serverless architecture leveraging AWS Lambda, DynamoDB, Elastic Search and S3 Leverage 3rd party apis for NLP and real-time messaging
Challenges we ran into
Conversational AI is a challenge, and requires lots of data to build sophisticated models.
Accomplishments that we're proud of
Slack integration for management and triage of calls.
What we learned
NLP is hard, and needs lots of data.
What's next for Amelie
Collect more data, so can build more intelligent platform.
Built With
- amazon-dynamodb
- aws-lambda
- elasticsearch
- node.js
Log in or sign up for Devpost to join the conversation.