CASE STUDY to improve socioeconomic empowerment. Technology has to translate into better outcomes for women entrepreneurs, not simply more systems.
You are developing an in-house AI platform. What business challenges is it designed to address, and why did you decide to build the platform internally rather than rely entirely on third-party solutions?
Our interest in AI comes from a very practical business need. WEF has a large and increasingly complex information environment. We have operational data, financial information, customer interactions, policies, procedures, reports and other institutional knowledge. The challenge is how we make that information more accessible and useful to the people who need it while maintaining appropriate controls.
The AI platform we are developing is therefore intended to address business problems such as information retrieval, knowledge management, operational analysis, management reporting, customer support and decision support.
For example, instead of an officer spending significant time searching through multiple documents, systems or reports to answer a routine institutional question, an appropriately governed AI capability should be able to help retrieve and synthesise the relevant information.
There is also an opportunity to use AI to identify patterns in operational and financial data. This could support areas such as portfolio monitoring, loan recovery, service demand analysis, anomaly identification and resource planning.
We are not approaching this as a decision to build everything ourselves. There are excellent commercial AI technologies available, and we will use them where appropriate. What we want to develop internally is the institutional intelligence layer, the governance, data
architecture, business rules, workflows, knowledge base and use cases that are specific to WEF.
There is a strong reason for that approach. WEF ' s data is closely connected to its mandate and operations, and we need to retain control over how sensitive information is accessed, processed and used.
We want to own the areas that differentiate WEF and use proven external technologies where they provide better economics, maturity or scalability.
What capabilities are you prioritising for the AI platform, and how do you expect it to improve operational efficiency, decision-making and the experience of the women entrepreneurs you serve?
The first is digital financial inclusion, expanding access to our services through channels that women already use and understand. The digital lending model delivered through mobile channels is an important part of that journey.
We are looking at AI from three perspectives: the employee, management and the woman entrepreneur.
For employees, one of the immediate opportunities is an intelligent institutional knowledge assistant. WEF has policies, procedures, guidelines, reports, manuals and operational information distributed across different areas. AI can help officers find and understand the information they need much faster.
The second area is analytics and decision support. We want to progressively move towards systems that can identify trends, exceptions and patterns rather than simply producing static reports. This could support management in areas such as loan performance, recovery, customer demand, operational workloads and resource allocation.
The third area is customer experience. In the longer term, AI can support more intelligent digital self-service, allowing women entrepreneurs to obtain information about products, applications, repayments, requirements and services through simpler conversational interfaces.
There is also an important opportunity around proactive service. Instead of waiting for a customer to ask a question or report a problem, digital systems should increasingly be able to identify where an intervention may be required.
But we are being deliberate about the role of AI. AI should augment human decision-making, not replace accountability. Where a decision has financial, regulatory or significant customer consequences, there must be appropriate human oversight.
Our ultimate measure of success will therefore not be the number of AI models we deploy. It will be whether an officer can do their work more efficiently, whether management has better information for decision-making and whether a woman entrepreneur can access WEF services with less friction.
How are you approaching the development of internal AI skills, data governance and responsible AI practices to ensure the platform can be scaled securely and sustainably?
AI is not primarily a technology challenge. It is also a people, data, governance and organisational-change challenge.
We therefore have to develop internal capability alongside the technology. Our approach is to build AI literacy across the organisation while developing deeper technical capabilities within ICT and the relevant business functions.
Different staff members will require different levels of capability. A business officer does not need to become a data scientist, but they need to understand how to use AI responsibly, verify outputs and recognise situations where AI should not be relied upon. Technical teams require deeper capabilities in areas such as data engineering, AI architecture, cybersecurity, model integration and system administration. www. intelligentcio. com
INTELLIGENT CIO AFRICA
15