While AI developments are proliferating around the world, there are several concerns around AI development in Africa. The hurdles to achieving rapid AI growth and development in Africa include a lack of sophisticated IT infrastructure for most of the continent, plus a lack of energy supply to power the large computational power required, and a shortage of skills.
The end result of this is that AI development in Africa, compared to global levels has been low, and predominantly concentrated in three regions only, that are; Nigeria, Kenya and South Africa.
The market size in the Artificial Intelligence market in Africa, is projected to reach US$4.92bn in 2025 compared to the United States alone where US$66.21bn is projected to be invested in AI developments in 2025.
Entry of an AI Challenger
Business tech Africa spoke to Ram Ramakrishnan, CEO of Cloud23 who is an IT architect by trade, and have just launched a brand new AI driven Chief Revenue Officer product called Fallon.
Cloud23 is a salesforce consulting partner in South Africa and throughout the rest of Africa where they work with businesses to transform business processes using Salesforce and other cloud solutions.
The company uses a product under licence within Salesforce called AgentForce, and this is the platform that they have developed Fallon on.
The AI CFO concept is within the Grouping, Pegging and Distribution (GPD) space by harnessing AI revolution to move it to beyond simply using AI for content generation, and producing image and articles.
“The true power of this tech is to interpret data, understand and advise on solutions or take a left, or a right kind of an approach” says Ram.
Budget constraints Sparked the Concept
He mentions what sparked the concept was when he visited an international conference and saw the way some big brands were using AI. “We started ideating, OK, how do we actually use it in our context in South Africa? Because we can’t really pump in hundreds of millions to get these things up and running” Ram reflects
“Then we distilled this down to our own internal problem to say, how do we automate at minimal cost so that we don’t hire people for jobs that are not interesting”? he comments.
This was the moment that launched the quest to seek out the solution and the best possible way to test this, which was, by creating a solution that worked for their own enterprise.
Ram and his team set out to use AI to enhance Cloud23 sales directly, by bringing all relevant sales data together, customer data, order data, invoices, purchase orders, sales successes and failures and focus on analysing these in detail all together.
“Obviously, we do projects” says Ram, “we manage services, and so we brought all the project data together, and when customers and projects are fed into a tool like AgentForce or an AI platform like AgentForce, then we can start interacting in a different way”, he says.
AI Agent Application
A typical use of the AI, Ram mentions is using Fallon in their normal processes. “We will chat with our bot or an agent on a Monday morning saying, how are things going on? Are we good for our quarterly target? Can we have a bit of an overview of what to prioritise this week?” he says “And now we started getting advice, and that sort of enablement and coaching is where we found the big gap for Fallon” he remarks.
Fallon is a massive jump up, especially for a start-ups and for scale-up businesses according to Ram. Businesses, like theirs, will see the great potential of having an effective AI agent that can peel back the layers without having to adding layers of more staff in order to achieve this.
For any Small enterprise to go and hire or recruit a CRO from a big enterprise, it comes with a price that is often out of the scope of what they can afford. “Then we kind of pivoted to this technology solution, says Ram, and the teams in our business are loving it.
Home Grown Advantage
The fact that Cloud23 has created a home grown product had many advantages for the business and they soon figured out how to use the AI to determine optimisations within the business by pairing the right person or offer to the right client or project and to optimise project approaches.
“So no more treating our consultants as typical consultants, but rather, we are making more insightful decisions and providing better insights to our people Ram states. “So those are things that provided positive outcomes, but we didn’t anticipate to enable such rapid tweaks and improvements” he says.
Multiple System Interface an advantage
Fallon is also able to interface with multiple systems and can import and utilise data from anywhere, including Xero, QuickBooks, Salesforce or any other CRM, billing and order systems and it works with any CRM for that matter.
Fallon can also create new leads based off of data that is provided to find suitable companies to target, but its major benefit is also identifying upselling opportunities to existing clients based on the history of sales seen with other clients with similar needs.
“More often, the immediate benefit that we’re seeing is on our existing customer base to say, where are the gaps? How do we fill the gap? Or before even filling, how do we identify these gaps? And what’s the right time to engage with the customer?” says Ram.
Fallon can automatically draft an email or create a proposal providing timely and targeted interactions with their clients.
The Price Tag
With major corporation spending millions on AI project investments, and not getting a return on investment, smaller businesses could be forgiven for not even considering an AI project at all.
Ram points out that that Salesforce has an entry level product that offers smaller businesses the opportunity to test their platform for free. This is however limited to around 100 conversations in a month.
However, if your business does not have anything in place it’s a starting point where people can engage, interact, understand how it’s going to work for them.
“And then you can either continue in that same level with minimum conversations, or you can top it up with more tokens, we call, in the GPT world” says Ram. “So that you’re looking at plus or minus, a bare minimum of around R40,000 to R45,000 (ZAR) (roughly $2175 – $2554) a year in tokens cost. And then roughly, even if you work with a 30% buffer for a two people team start-up, you’re not looking more than 80,000 per year”, He states.
Data Structure Key Aspect of AI
The key thing to determine for any company wanting to deploy an AI agent in their business is the data and the data structure. There’s no point in deploying an agent without the data.
Cloud23 will consult with companies and determine how and in what shape their data should be brought together, to provide the best impact.
Small companies may not have the massive data sources that large corporates do over hundreds of systems. Perhaps only a few, so the project would be based on the complexity of information that they provide. “The beauty about Fallon”, says Ram, “is that it also takes unstructured data”. This type of data is often very useful in determining and assisting in areas within a company such as HR and human development.
Ram believes that the best approach for an AI agent such as Fallon is to have a ROI-based approach to outline what companies want to achieve and then work backwards from that, rather than generically saying we want an X AI system.
“It’s very important to actually scope something and to say what are we hoping to achieve first rather than the other way around” he concludes.
Read More About AI in Africa- READ HERE
