Inspiration
IN South Africa we are a mixed society where there are many classes and races. The different backgrounds coupled with our history makes us a uniquely unequal society. This has led to many people falling outside of the formal banking sector. This is further exacerbated by the systemic inequality that plagues our country. This has led to a concentration of banking services that are not tailored to the small business owner or low pay individual. This situation has been ameliorated by the advent of new banking institutions such as Capitec, African Bank and Tyme bank all of who are looking to cater for the unbanked and underbanked. The ease with which accounts can be opened and the network of branches has made the process less strenuous for new accounts to be opened. The banking sector does still remain one of the most untransformed and unequal in the country.
In this hackathon we have seen there is a multitude of solutions that can enable those who are still underbanked to receive services that are held within the domain of the formal banking sector. The same ease can be further simplified by reaching people on platforms that they are accustomed to. The use of WhatsApp for an African market makes sense as this is and remains the most widely used social interaction point for most Africans as it crosses many barriers and technological skillsets. Placing a chatbot on this service allows us to have access to the many millions of Africans across the continent who are not able to access the formalised banking sector. This also allows for transactions to happen across the African continent and globally.
What it does
The chatbot interacts with people who chose to have it to enable them to start their savings journey and also to manage their money matters. The chatbot acts as an onboarding mechanism for people who don't already use the banking system to find means and alternative ways of saving. The main aim of the chatbot is to get people who don't bank with any particular bank to start banking.
The chatbot engages customers on what they need the money that they make for. Then it assesses if the user is using the money for a specific target or goal. The chatbot then assists them to meet their target. Basically it will advise them on Finastra products that will assist them in making their targets.
The chatbot would also be able to make and receive payments into an account that if the user wants would be created. We don't want to be another cashapp or paypal but we want them to have the ease of integrating those payment reception platforms into one unified platform where they can access savings interest rates that are competitive and profit making.
The chatbot is almost a financial coach that would be trained to identify spend patterns that fall out of line with the individual's targets. Targets are also very simple "saving to pay rent" "spend no more than $25 a week on groceries". These simple targets are then the daily interaction that the chatbot will talk about.
The chatbot then will send messages that encourage and focus the targets for the user. Something simple like "Today you saved $2 by taking the early bus", "You are about to start spending $% more than your target for this week. Manage your budget accordingly."
the chatbot is able to post consumer driven transactions without the need to log in to an alternative platform. So a user does not need to go into their banking app to transact and make payments. The security features of 2 factor authentication and end to end encryption mean that the conversation is secure and can only accessed by the user of the phone.
How we built it
The chatbot was built using the python flask framework for it's lightweight nature and used the rasa NLU package to create the conversation and actions the the chatbot would need to function. The chatbot is able to make note of dates and send messages to the user about prerecorded dates that are of a financial nature. The chatbot is able to do basic calculations of money which we built using the built in python packages. The chatbot is also able to assess a user's spend relative to their targets by means of the conversation between the user and the bot. The analysis of the spend patterns are aligned to what the user inputs and as such is not always entirely accurate if the user does not enter in the details truthfully. The chatbot has intents that are linked to certain services that finastra provides through their API directory. The api's that we will be building into the chatbot are: Remote Deposit Capture Bill Payments Alerts Branch Remittance ATM and POS Transactions Current and Savings Account Onboarding Customer Onboarding Internal Transfers Good funds Payments Person to Person payments
Challenges we ran into
We started the Hackathon late and as such we were a little under the gun as we were not able to gain access to FusionFabric in too much depth. We also found the lack of mentors to be a particularly challenging aspect of the hackathon as we were not able to get insights on the inner workings of the API's and service offerings. We also are not sure of Finastra's growth path and whether Africa features on the agenda at all. Our biggest challenge is our limited knowledge of the US banking system. Google can only tell you so much.
Accomplishments that we're proud of
We were able to build a working chatbot that was able to send API requests to the Fusion Fabric API's. With this chatbot we could onboard users and get new users into the formal banking system without the need for users visiting banks. We could also enable any small business owner or individual to start investing and saving their hard earned money in a way that allows them to generate positive financial attitudes.
What we learned
The banking systems of the world are not uniform and the red tape exists for a lot of very good reasons. The problem is that this red tape is also a hinderance for many legitimate users who are unable to meet the requirements. Banking could happen remotely and without the need for physical structures.
What's next for Zaka
Plug in to the local banking sector or fintech sector as there are a great many services that could benefit with the use of technological innovation. Build a string product offering for an African audience across social media platforms. Increase the security of the transactions and the platform for data security reasons.
Log in or sign up for Devpost to join the conversation.