Technology has a remarkable potential to do good and
solve real-world problems – and innovations in areas like AI are proving it.
Let’s Talk About Sex and Open Kasi, for instance, are two of the latest
solutions designed to help make a positive change in financial services and
health in South Africa.
Let’s Talk About Sex (LTAS) is an AI-powered chatbot
aimed at providing critical sex education and support to South Africa’s youth
around issues such as HIV, sexually transmitted diseases (STDs), unplanned
pregnancies and abortions, as well as around broader concerns such as low
self-esteem and depression. It uses machine learning to answer any questions
users may have, and carries out a sentiment analysis on all questions to help
develop relevant and empathetic answers.
Open Kasi is geared towards helping small business
owners in townships streamline their financial functions and documentation such
as invoices so that they have a real-time, any-time view of how their business
is performing. It provides vital insights that allows these entrepreneurs to
start, run and manage their businesses even without an in-depth understanding
of fintech or financial aspects such as invoicing.
The teams conceptualised and developed the prototypes
of these life-changing solutions at Microsoft’s AI for the Financial Services
Industry (FSI) and Health hackathon in June. The event brought together talent
from varying disciplines to collaborate on creating unique solutions for real
problems, particularly affecting the financial services and health industries,
using AI and machine learning services in the Azure Cloud.
LTAS and Open Kasi emerged as the winners from the six
teams chosen to participate. “There was significant interest in the topics – we
had to narrow it down to six teams from 39 proposals – because these sectors
are always top of mind and there is a massive potential for impact by providing
solutions to real-life issues,” says Asif Valley, National Technology Officer
at Microsoft South Africa.
Building for future needs through diversity and
inclusion
As part of creating tangible impact, these solutions
broadly focused on building diversity and inclusion into them – and the
diversity of the teams involved in the hack ensured that the solutions have the
ability to cater to the needs of South Africans across the spectrum inherently
built into them. This speaks directly to the hackathon’s aim of innovating and creating
for good.
Another key aim of the hackathon was giving teams the
freedom to explore their own ideas of what solutions these industries needed
most. Each team got to define their own interpretation of meaningful change and
create a unique solution to drive that change. “We wanted to get fresh
perspectives that might leapfrog traditional issues and provide solutions that
we might not even have thought of that build for future needs,” says Valley.
There were, however, guidelines that teams needed to
work within. Teams could, for example, use any operating system, development
language, framework, tools and hardware – and while solutions can stretch
across multiple cloud platforms, all machine learning or cognitive technology
aspects of the solution must have used Microsoft Azure Machine Learning or
Microsoft Azure Cognitive Services.
Although the Open Kasi and LTAS solutions are still
just prototypes, they are now one step closer to becoming a reality. The teams
will receive mentorship, guidance and training through Microsoft’s Head Start
programme. Head Start provides access to digital courses and training, live
events, meetings and mentorship, which help take the prototypes to the next
level and make them more commercially viable.
“Ultimately, we’d like to see these solutions in
action so we can see the concept of using AI for good in action,” says Valley.
“The AI for FSI and Health hackathon truly showed the potential of AI to solve
real-world problems, so we are looking to grow it to include other sectors such
as agriculture as well as get key stakeholders involved so we can focus on
collectively developing the principles and boundaries that govern the ethics of
AI.”
