South Africa

Botlhale AI

  • Company representative: Thapelo Nthite, Xolisani Nkwentsha, Sange Maxaku
  • Data enthusiast: Derwin Ngomane
  • Data science expert: Cobus Burger
  • Problem statement

    An investigation into the factors impacting financial exclusion in South Africa has shown that Language is one of the most significant factors. Customer engagement is a key factor in driving under-served, low-income individuals to gain access to financial services and, more importantly, to use those services effectively. IsiXhosa and IsiZulu speakers are the largest groups in South Africa making up 16% and 22,7% respectively. Yet they have the lowest probability of having a bank account with only 30% and 26% of them having a bank account. Non-English speakers makeup more than 80% of the South Africa population yet banks offer most of their services in English thereby excluding most of the South African population.

    Solution description

    Our solution is a multilingual chatbot for banking services. It allows users to connect directly to their banks and perform banking transactions using natural language, through text and voice messages. The user simply writes to the chatbot the transaction they wish to make and the chatbot will process the transaction (like transferring money, buying airtime or electricity etc.). The chatbot then understands the intent of the message and performs the relevant transaction and response. We are helping banks and their customers in four ways: (1) We are simplifying banking for their customers by enabling them to interact with banks in simple natural language conversations. And by having these conversations where their customers already are, on platforms like WhatsApp and Facebook messenger. (2) Secondly, we help banks build more personal relationships with their customers beyond just transactional banking. By proactively helping their customers plan and take action to improve their financial lives. (3) Thirdly we reduce the banks running costs by limiting the amount of human interaction required. (4) Lastly and more importantly, we make banking more inclusive by interacting with in languages they trust and understand.

    Market segment

    Our target is a subset of the 80% of the South African population that doesn’t speak English as their first language and are not comfortable with the language. More specifically people who use mobile banking services or people who could be using mobile banking services if they were accommodated. There is a continued growth in the number of people and Financial Institutions who engage through messaging platforms like WhatsApp. The ideal customer is people who are not comfortable with English but wish to engage with banks through their mobiles. We recently did a survey of non-English speakers in Cape Town, South Africa (which has mostly Xhosa speaking people) that showed that people would want to use our solution and their preferred chatting platform is WhatsApp.