Inspiration
As technology evolves, more people are being left out due to some reasons which include having low digital literacy and exclusive technological solutions.
What it does
This app is built for people having low digital literacy, unable to communicate through typing, and those who have difficulty carrying out tasks on their own. They can get help from volunteers to complete their tasks/chat with them online. With a few simple taps, they will be able to send out tasks/chat requests to the platform. An AI model is used to generate tags from the transcribed live recording and separate their messages into different categories. Volunteers will then be able to pick tasks based on their availabilities and preferences based on the different categories.
How we built it
- Web app - Django & React
- AI logic - Python, AutoML
- APIs used - Deepgram (features : keyword, interim, language)
Challenges we ran into
- Training the model using Google Cloud AutoML (takes too long)
- Finding a suitable dataset for the ML model. We then set our own rules to generate suitable tags and collect data where tags couldn't be determined so ML can be implemented in future.
- Extracting data from the database
Accomplishments that we're proud of
Managed to build a working prototype that includes live transcription using the Deepgram API.
What we learned
- Learned to build an app and integrate the API
- Work under time constraints
What's next for Lend your Hands
- Include user authentication and collect more data to tailor their needs and find suitable volunteers. (location, special needs)
- Add the language option to overcome language barriers between immigrants and locals
- Add the chatbox function for tags like emotion and entertainment
- Implement better AI model to predict user's needs (especially those who have speech disorders)
Log in or sign up for Devpost to join the conversation.