15 million adolescent girls aged 15–19 have experienced forced sex in their lifetime worldwide. Roughly 3 in 4 children between the ages of 2 and 4 – around 300 million – are regularly subjected to violent discipline by their caregivers.

What it does

Tambua is a Swahili name meaning "realize" (in the English language). The project raises awareness about child violence through facts about child violence and increases the level of understanding related to child violence.

How I built it

I used RASA technology to build an AI conversational chatbot that responds to user queries. The brain of the chatbot is on Heroku. The UI communicates with the brain through an API.

Challenges I ran into

  • The chatbot is implemented mainly in the Swahili language rather than in English
  • The first response of the chatbot can take a couple of minutes due to the status of the Heroku server

Accomplishments that I'm proud of

  • The chatbot responses are 88% accurate to user queries
  • The chatbot is 24/7 available
  • The uptime of the server is 99.999%

What I learned

  • I learned about RASA technology
  • I learned about machine learning, supervised embeddings, and natural language processing
  • I learned about integration between backend and front-end

What's next for Tambua

  • To train the chatbot to be at least 98.9% accurate

Built With

Share this project: