Joyful Helper

AI assistant that empowers migrant domestic workers

in submission for WWcode hackathon for social good 2023

Challenge Statement

1. Globally, there are approximately 169 million migrant workers, constituting 4.9% of the global labour force in their destination countries. Many of these workers come from countries with a different native language.
2. These workers often face language and digital literacy challenges in accessing essential information, such as worker rights, in turn resulting in labour exploitation.
3. Hence, we aim to use Generative AI (LLM) to bridge this gap by providing timely and accurate information in their native language via a messaging platform (Facebook Messenger) they are most familiar with.

Solutioning

Development:

• Python as primary language
• Iterative prompt engineering of openAI GPT-3.5-turbo model to optimise contextual understanding

Deployment:

• Facebook Messenger integration with Messagebird
• Flask app hosted on pythonanywhere
• SQLite for context memory

User engagement:

• Immediate Response
• Origin and Destination country-specific
• Language-specific

Try it at

Messenger link to Joyful Helper

Team

by team LLMpowerers: Sandra Tan, Winnie Low, Jin Jin
with dev advice from: Yong Sheng