Africa, AI and Worktech: Visiting Workpay in Kenya.
I have a cool job. Last week I headed over to Kenya for a board meeting and strategy weekend withWorkpay, one of our portfolio companies. I got to spend quite a bit of time with the board and leadership team,. . .
I have a cool job. Last week I headed over to Kenya for a board meeting and strategy weekend withWorkpay, one of our portfolio companies. I got to spend quite a bit of time with the board and leadership team, participate in a panel discussion about AI and work, join a podcast, and later bore the most of company with my anecdotes about how payroll software works. I also was able to see a bit of lovely countryside, the city and an art museum with my son, we also had a night time hippos-on-land viewing and a hike up and around a dormant volcano called Mount Longonot. Not all board meetings are like this, alas. Serendipitously just before the board meeting this image popped up in my news feed (I suspect I’m one of a very small number of VCs that uses the Guardian as a primary news source). More details on the Guardianstory here. On the one hand, this is shoddy work from the State Department. The human in the loop was AWOL, and they quickly apologized. But more significantly it neatly highlights the weaknesses of US foundation models when it comes to African data and context. This is not simply a hallucination, it is ignorance as code. This is a robust metaphor for why the AI and software solutions to solve Africa’s challenges will not be merely bestowed from abroad. Sure, Africa will leverage international capabilities, but it will take local ingenuity, research, context and leadership to make it work for the benefit of Africans. On the first day of my trip, I was lucky enough to be on a panel with CEO ofQhala,Dr Shikoh Gitau. Amongst many other impressive things, both academically and commercially, she has founded an innovation lab focused on building and applying AI to African challenges. She combines deep AI technical chops with a profound understanding of the African context for AI. She spoke knowledgeably and passionately about applying AI to the unique challenges of medical care and legal tech in Africa, see for example theAfrican pathology networkandThe Center for Epidemiological Modelling and Analysis (CEMA)and in legal tech,Haki. Africa’s problem context is so far removed from that of Silicon Valley, it is no wonder that many of the solutions posited by Silicon Valley fail here. Silicon Valley is not the Rift Valley. I rather like this quote from Shikoh’s feed. Her belief is that the most transformative AI solutions for Africa will come from Africa. While I have you, let’s remember that many of the large AI providers have taken advantage of African labour to train their AI. The models we use have a real human cost, and a chunk of that is borne by African workers. As Mark Grahamnotes in his latest book. Silicon Valley has sold us the illusion that artificial intelligence is a frictionless technology that will bring wealth and prosperity to humanity. But hidden beneath this smooth surface lies the grim reality of a precarious global workforce of millions labouring under often appalling conditions to make AI possible. This book presents an urgent, riveting investigation of the intricate network that maintains this exploitative system, revealing the untold truth of AI. It is excellent to see the exploited workers arebeginning to fight back. Techno-colonialism is a thing. Technology is not neutral. Demographically speaking, Sub-Saharan Africa is expected to account for 90% of global population growth over the next 25 years. This means that many more companies will be employing people in and from Africa. There is massive work to do in building job opportunities for those people, but it is reasonable to assume that the population at work in Africa will grow dramatically. As the formal sector grows, so does the demand for business tools such as payroll. The growth for African payroll technology is inevitable. Yet when international software and service companies talk of “global payroll”, I often wonder if they have ever set foot in the countries they purport to support. Almost always, they have not. The term global payroll invokes an image of an automated engine that magically processes the world’s payroll, but the reality is far more prosaic. Typically, behind the trivial facade of a “global” dashboard, a file is emailed to a local provider, who calculates the payroll, often with a spreadsheet. That information is then emailed back and forth. Tax filing processes differ vastly between African countries: Some have modern digital processes that are more advanced than the US and much of Europe, others need a person to stand in a long queue to manually file documentation. Payment rails are often inefficient and poorly audited. Responsibility and accountability are hard to pin down. US and European vendors bragging about setting up a few local entities and calling it full employee of record (EOR) irks me more than a bit. Across the Acadian Ventures portfolio, we have seen several innovative uses of AI in a payroll context, whether it is within the payroll engine itself, in the legal change process, data management, service delivery, employee pay communication, early wage access, EOR, tax processing, stablecoin, financial services integration and more. I spent quite a bit of time with the technical and product folks at Workpay (Ajaykumar and David) and their product stack compares well against their peers from Europe and the US. For instance, the work that they are doing with payroll wallets and financial services integration is more innovative than I have seen anywhere else. The core payroll engine now processes full taxation calculations for more than 40 African countries. Their employer of record offering goes way beyond mere legal entity provision, and has robust workflow automations, separation of duties, native1gross to net, and where possible, automated government electronic filing and payment rails. They have by far the most robust African payroll regulatory context model for building robust AI payroll capabilities. They are also working on deep integrations into incumbent platforms for multinationals (Think SAP and Workday etc). Workpay also benefits from a strong of African focused investors. They have helped scale other successful African companies, so I’m learning a lot fromRio,Ngetha,Maina, andAliabout scaling African companies and the broader VC environment and ecosystem. The founders, Paul and Jackson, are determined to build an African powerhouse and they have built a strong team. It was great meeting Boris, Ferdinand, Ajaykumar, Jess, David, Ernestine, Brian, Jacqueline, Ann, and so many other members of the team in the office and at the offsite. Where other companies just have pot plants, they have a veggie garden. Board meetings, podcast and office. Workpay is stepping up its partnering with international EOR and payroll providers, as well as winning local and international enterprise deals directly. Here’s a case studyfrom Silensec, a multinational cybersecurity company andanother onefrom Global Health Corps, andone from ATP. I rather liked this quote from the ATP MD, Sarah McNeely. What I wantedwas first and foremost an afrocentric solution. I wanted an African company to help us build our African company. We wanted to hire people in Nairobi, in South Africa. We wanted to hire contractors all across the continent from Zambia to Malawi toCote d’Ivoireto Senegal to Kenya and I needed someone who had that coverage. Also the University of Toronto manages their African research project workforce in Kenya, Ghana and South Africa with Workpay. There are powerful solutions built by people on the continent that can stand up technically to their international counterparts, yet massively outperform them through rich local support, service and requirements understanding. Whether you are a global organization seeking a reliable partner for EOR or local entity payroll, or an African headquartered company looking to manage your workforce more effectively across the continent, I suggest you take a good look at Workpay, dig under the hood. Then ask the “globals” about Tanzania’s tax filing rules. To the regulators, Workpay provides a robust and efficient way to collect employment taxes, reducing errors and fraud. I’d urge you to work with companies like Workpay to open up secure APIs. Electronic payroll is a building block for financial inclusion. For those “global” players, you would be better off partnering with Workpay and focus on what you are best at. You may also find my other posts about Workpay and “global” payroll useful. John Odoch-Ameny sculpture. While I could share many lovely photos of the Kenyan countryside and its wildlife, I thought you’d like this sculpture from John Odoch-Ameny. His work asks some awkward questions about culture, technology and society. It is at the Heritage house, just outside the Nairobi National Park. If you are interested in African Art and Culture the house is a must see. It is marvelous. Also there is something special about seeing an Acacia tree. IYKYK. More generally, if you are planning a Kenya trip, I’m happy to connect you with the folks that made our visit so memorable. As I almost always do I will end with a tune. For this week, I’m heading further South, to where I grew up. The mighty Brenda Fassie. I defy you to sit still. 1I find using the term native really awkward here. It has a most unwelcome legacy in an African context. No posts