Why enabling and profiling African data and models matters? has been saved
Why enabling and profiling African data and models matters?
Deloitte’s trusted partnership approach is designed to empower African Data scientists for African insights
"Bringing Africa’s Data Scientists together around the ethical use of data insights will ensure that consumers across the continent are better served and, will enrich global business offerings with African based insights and data. This ultimately creates commercial value at an enterprise level and social impact beyond the footprint of your company." – Dr. Quentin Williams
“Organisations globally, including Deloitte, are transitioning their advanced analytics and artificial intelligence initiatives onto the cloud. The ability to access scalable, high end compute capability, combined with access to native applications and algorithms, is a compelling value proposition. The fact that it is easy for practitioners to spin up an environment, burst capacity and only pay for it while using it creates a new level of freedom and flexibility. However, taking your organisation’s cloud strategy for analytics to enterprise level requires a myriad of additional factors to be considered. To mention only a few, new ways of work, streamlining your operating model, enterprise wide governance, information security and access management, network performance and integration, deployment and billing controls. These are some of the aspects that prove to be more complex than anticipated and often underestimated. Deloitte has assisted clients with some of the largest data and analytics transformation programs in Africa. Through this we have gained valuable practical experience to ensure that your cloud enabled journey towards an Insight Driven Organisation delivers tangible value, making a real impact where it matters.” – Werner Swanepoel
Deloitte’s recently published Tech Trends 2020 highlights how every aspect of a company that is enabled by technology “represents an opportunity to lose – or earn – trust with customers, employees, partners, investors, and/or regulators”. This is often related to the level of trust in the data; understanding the outputs from data science models and the ethical use of insights from these models. Context relevant, particularly for the African continent, and transparent Artificial Intelligence (AI) therefore becomes a business-critical goal across the many dimensions of an organisation’s technology, processes, and people
Dr Quentin Williams, leader of Deloitte’s Consulting Data Science practice believes that for the African consumer, this is a key consideration. Decisions made around their interactions – for example credit approvals, product recommendations and content searches – are often dependent on models trained on data from elsewhere in the world. This leads to algorithmic bias and mistrust in your company. Anyone from the continent who has ever asked Siri or Alexa for the meaning of their name in their language, understands that the current global data universe has a shortage of African-based data and insights.
On the other side “businesses have been asking customers to trust them in new and deeper ways”, from asking for personal information to basing customer insights on tracking online behaviour. This must be done by methodically and consistently pivoting algorithms to collect and analyse data, ethically, based on circumstance. For example, design a credit score locally, using models based on the nuances of your local context, and not designed by external parties based on data collected from external groups of customers.
In this light it is important for companies on this continent to ensure that a strong data science community is created within their organisation and with their ecosystem partners that:
- Creates a treasure trove of African data and models that can be shared and re-used
- Ensures customer-facing product solutions are built on data and models relevant to them – an almost data “jury of peers” – and infused with global insights through symbiotic transfer learning
- Is enabled by cloud-based technology that drives efficiency, speed to production and removes friction to value.
The use cases presented in Deloitte’s Tech Trend 2020 report: Google’s internal security turnaround; Providence St. Joseph Health standardisation of cloud and AI platforms; and CIBC’s organisation-wide AI enablement, practically shows that removing bias from data and empowering the data science community is a 360-degree opportunity. This ensures that the many dimensions across an organisation’s technology, processes, and people are working in alignment to maintain the high level of trust expected by their many stakeholders
For Deloitte, currently running two of the largest data and analytics
transformational programmes on the continent, this is an important consideration. For us taking your Cloud analytics usage to enterprise level means that a myriad of additional factors needs to be catered for. This include enterprise-wide governance, data compliance, data security, data quality and data governance, cross-border integration and deployment, and the ethical use of analytics based on value. This is becoming even more of a burning platform towards ensuring customer data is secure and relevant to their context, given recent data regulations published across the continent (e.g. South Africa’s POPIA act). An integrated AI cloud platform with on-prem and cloud capabilities ensures that data regulations can be encoded and measured within technology while acting as a value driver.
Deloitte’s approach is therefore designed to empower African data scientists with the right technology, appropriately constituted, for them to create African-based, respectful and relevant consumer product solutions; and most importantly monetise and externalise their models for their consumers and the global community.