👨🏿🚀TechCabal Daily – Sunset at Gigbanc
In partnership with Lire en Français اقرأ هذا باللغة العربية Good morning. Welcome to another capitalism-filled work week. I’m (begrudgingly) on lede duty all week, so you’re stuck with me until Friday. Since you have no choice but to give me your attention, here’s a question for you: We’re near the end of the 2026 World Cup, and despite not being much of a football fan myself, I can’t escape the debates. So, who are you backing to lift the trophy (no, you can’t change your answer after the final whistle)? Before someone starts another ‘Messi vs everybody else’ debate, here’s today’s dispatch. — Yemi Get smarter about Francophone Africa with our newsletter, Francophone Weekly—the startups, tech policies, and institutions building the pipelines for ecosystem growth. Subscribe Gigbanc winds down operations SA wants devices that check a foreigner’s status KRA posts record tax collection for FY2026 Portugal and Morocco want to build a subsea electricity cable World Wide Web 3 Job Openings Startups Nigerian fintech Gigbanc winds down its operations Gigbanc team If Africa’s tech ecosystem had a town crier, it would echo what the players already know: investors are becoming more selective. Startups are finding that building a product is one challenge, but raising enough money to keep building it is another. What happened? Gigbanc, a Nigerian fintech that built cross-border payment products for freelancers, is winding down its operations after three years. Customers have until July 31 to convert their balances to Naira and withdraw non-fraudulent funds to Nigerian bank accounts free of charge. The startup also said that it is in talks to be acquired. The ecosystem has money problems: Gigbanc’s closure comes at a curious time for African startups. Funding into the continent ticked up slightly in the first half of 2026, but the number of disclosed deals fell sharply from 252 in H1 2025 to 146 this year. Meaning that investors are still writing cheques, but they’re writing them for fewer companies. Where Gigbanc fits in: Gigbanc said it struggled to raise fresh capital while operating a business that is expensive to run. Cross-border payments require customer verification (KYC), compliance, banking relationships, and payment infrastructure, all of which cost money even before a startup becomes profitable. The startup tried to change tack away from cross-border fintech operations, but didn’t secure enough money to fund that ambition. Like Chimoney, another Nigerian-founded cross-border payments startup, that shut down in May after its founder said the startup could not raise additional funding. Early-stage startups need capital to build products, acquire customers, prove their business works, and reach the scale investors want to see. But investors now want to see the scale before writing the cheque. That leaves many founders stuck in the middle. Zoom out: Gigbanc said it processed more than ₦10 billion ($7.2 million) in payments, served 150,000 users across more than 30 countries, and built products for Africa’s growing freelance economy. Yet, even that wasn’t enough to guarantee survival. Modern Rails for Africa’s Economy: How Fincra is helping businesses collect, pay out, convert, and settle across African markets. Read more here. Policy South Africa wants to introduce handheld biometric scanners to identify undocumented migrants DHA minister Leon Schreiber. Image Source: Polity SA If you’ve ever gone through passport control at an airport, you’ve probably seen an officer scan your passport, take your fingerprints, and know almost immediately whether you should be allowed through. South Africa wants that same experience to happen on the street. What’s happening? South Africa’s Department of Home Affairs (DHA), the government department responsible for immigration, citizenship, and civil registration, wants to buy 600 handheld devices that will let immigration officials scan fingerprints and faces, then check in real time whether someone is legally in the country. Instead of relying on passports or permits that could be forged, officials would verify a person’s biometrics directly against the DHA’s database. Explain like I’m new here: The idea isn’t entirely new. Since April 2024, the DHA has been live-scanning people arrested on immigration grounds into its Automated Fingerprint Identification System (AFIS), allowing officials to identify repeat offenders and people previously deported. The new handheld devices simply take that same system into the field, so officers can verify someone’s status within seconds instead of taking them to an office first. Today, immigration enforcement often depends on physical documents and follow-up checks that can take time. The proposed system shifts the focus from documents to identity itself. If your fingerprints or face already exist in the DHA’s records, officials can immediately confirm your immigration status. If they don’t, your biometrics can be captured and stored for future enforcement. Why now? Immigration has become one of South Africa’s most politically charged issues. The DHA says deportations have increased—over 53,000 people had been processed for deportation or voluntary repatriation by July 11—as it invests heavily in border technology, while anti-immigration groups continue to pressure the government to crack down on undocumented migrants. Faster biometric checks are meant to make enforcement quicker and reduce reliance on physical documents that can be forged or borrowed. The catch: Buying scanners is easier than fixing the immigration system behind them. The technology assumes the DHA’s records are accurate, up to date, and accessible wherever officials are conducting operations. Mistakes, outdated records, or connectivity problems could leave people wrongly questioned or detained. Even if the system works perfectly, identifying more undocumented migrants only helps if detention centres, tribunals, and deportation processes can keep pace. Zoom out: South Africa is moving immigration enforcement closer to real-time policing, where identity can be verified almost instantly. Whether that makes the system fairer or simply faster will depend less on the scanners than on the institutions behind them. Naira Life 2026 is here! The theme for this year’s Naira Life Conference by Zikoko is “All About Wealth.”Join 2,000+ in Lagos on August 22 for a day of practical money conversations and workshops designed to move you from simply earning an income to building lasting wealth. Get 15%
Read More
The Next Wave: The million-dollar asset that no one can buy
Cet article est aussi disponible en français <!– In partnership with –> First published on 12 July, 2026 For more than a decade, venture capitalists and founders have clung to the belief that if a high-growth startup runs out of cash, its proprietary technology will retain enough value to soften the blow. An administrator can sell the company’s code, platform, or data to a strategic buyer and recover at least part of the investment. It is also often wrong. When a tech startup fails, its assets are worth only what a buyer can legally use. Delivery trucks depreciate. Custom software can become a liability. If a company’s data practices, licences or regulatory compliance are flawed, even technology developed at significant cost may become unsaleable. In insolvency, regulatory compliance, rather than intellectual property, often determines whether any value remains. Why data privacy is the ultimate gatekeeper Every consumer-facing technology company regards its customer database as one of its most valuable assets. That assumption has shaped startup valuations for years, as user data is seen as the foundation for future revenue. When United States retailer RadioShack filed for bankruptcy in 2015, its database of 65 million customer profiles was widely regarded as one of the few assets with significant value. Without it, the brand itself was considerably less attractive. A decade later, the collapse of genetic testing company 23andMe exposed the limits of that assumption. Its database contained genetic and health records that could not simply be sold alongside the rest of the business. More than 25 US state attorneys general, together with the US Trustee Program intervened, arguing that any transfer of such sensitive information required close judicial scrutiny. The court appointed a Consumer Privacy Ombudsman to assess whether any proposed sale would honour the privacy commitments made to customers. The same legal tension exists in Kenya under the Data Protection Act, 2019. The Office of the Data Protection Commissioner (ODPC) has progressively narrowed the circumstances in which organisations may rely on broad or implied consent when collecting and sharing personal data. Decisions such as Nancy Wansato Maroa v Vivo Energy and Artcaffe have established that organisations must clearly explain why personal data is being collected and identify any third parties with whom it may be shared. Consent that does not meet those standards may be deemed invalid. For an insolvency practitioner, those rules can turn a seemingly valuable customer database into an unusable asset. A startup that promised customers it would never sell their personal information cannot simply abandon that commitment because it has run out of money. Any attempt to transfer the database may expose the company and its directors to regulatory sanctions while leaving the buyer unable to lawfully use the data. The result is a striking inversion of conventional venture capital logic. The database may still exist, but without the legal right to transfer and use it, much of its commercial value is lost. Get smarter about Francophone Africa with our newsletter, Francophone Weekly—the startups, tech policies, and institutions building the pipelines for ecosystem growth. Subscribe For digital lending platforms and other non-deposit-taking credit providers, the principal asset on the balance sheet is the loan book: the portfolio of outstanding loans and receivables. Conventional financial thinking suggests that a distressed loan book is a liquid asset that can be sold or assigned to a commercial bank or a specialised debt collector. In fintech, however, regulatory compliance is the determining factor in whether those assets are enforceable. Under Section 33S of the Central Bank of Kenya (CBK) Act, operating a non-deposit-taking credit business without a valid CBK licence is a criminal offence. Next Wave continues after this ad. What happens when investors, visionary founders, policymakers, enterprise leaders, and innovators share the same room? ForgeTech Summit 2026 is bringing together leaders shaping the future of technology, capital, policy, and innovation across Africa and beyond. Designed as a highly curated gathering, ForgeTech creates the environment for meaningful conversations, strategic partnerships, and opportunities that extend well beyond a single day. 31 July 2026 | Nairobi, Kenya Request an invitation. The consequences of non-compliance were illustrated in the landmark case M-Collect Limited v Mbana Kalua, in which the High Court of Kenya dismissed 139 debt recovery suits brought by digital lenders. The court held that unlicenced lenders lacked legal standing to recover debt, collect outstanding sums or exercise statutory powers of sale against defaulting borrowers. Consider a digital lender entering administration with a loan book valued at KES 500 million ($3.9 million). If that company operated without a valid CBK digital credit provider licence, the portfolio’s recoverable value may effectively be zero because the debts cannot be legally enforced. Furthermore, under the Business Laws (Amendment) Act, an unlicenced entity may face statutory fines of up to KES 20 million ($155,000) or three times the financial gain derived from non-compliance. What appeared to be a valuable asset can quickly become a legal liability. Even if an administrator attempts to assign the debt, the absence of a registered financing statement on the electronic registry established under the Movable Property Security Rights Act (MPSR) may leave the security interest unperfected and ineffective against competing creditors. The case of climate tech The collapse of Koko Networks in January 2026 offers a cautionary case study for the climate-tech and clean-cooking sectors. Over more than a decade, Koko invested an undisclosed sum in developing a carbon-financed bioethanol cooking network serving 1.5 million Kenyan households. Its business model relied on selling subsidised smart cookstoves and fuel while generating certified emissions reductions that could be monetised as carbon credits in international carbon markets. That model unravelled after the Kenyan government declined to issue the Letter of Authorisation (LoA) required under Article 6 of the Paris Agreement for the relevant cross-border carbon credit transactions. Government officials, concerned about preserving the nation’s domestic carbon budget and reacting to academic critiques of cookstove emission methodologies, refused to greenlight Koko’s cross-border credit transfers. Without the required sovereign authorisation, Koko’s carbon-credit revenue stream ceased,
Read More
“We weren’t really sure what was going to finally stick”: Day 1-1000 of Rayda
Ask Francis Osifo, co-founder and chief executive officer of Rayda, a fixed-asset and information technology (IT) device lifecycle management platform, where the company began, and he will tell you a story about a laptop. Osifo was chief technology officer and a co-founder at 54gene, the genomics startup that would later shut down in 2023. The company had hired an employee in Kenya and needed to ship a laptop from Nigeria, he recalled. The device eventually spent nearly two months stuck in customs, and clearing it had become an expensive ordeal for 54gene. However, the problem that he noticed was bigger than one laptop. He explained that he had no reliable way of knowing what assets 54gene owned. The company’s finance records could show the value of its fixed assets, but they could not tell which employee currently had which company-issued device or whether it had been repaired or reassigned to someone else. As 54gene grew past 200 employees, that knowledge gap had started looking like a structural blind spot. “I kept going into that rabbit hole of the different parts of the problem, whether it is visibility around the data tracking, the depreciation of the value of those assets over time, or even being able to deal with that entire journey from an employee joining a company to an employee leaving the company,” he told TechCabal in an interview on Tuesday. Osifo left 54gene in 2022, convinced there had to be a better way to help businesses manage their IT assets. In January 2023, he launched Rayda, an IT operations platform that helps companies procure, deploy, manage and recover employee devices. A human resources or IT manager could request a device when a person is hired, and from there, Rayda’s system tracks it through procurement, shipping, storage, repairs, and retrieval when the employee leaves. The startup has since evolved from an asset management tool into what Osifo describes as an ‘IT command centre for globally distributed teams’. Rayda operates in the device lifecycle management industry, estimated at $4.8 billion. It also operates in the category of startups building infrastructure for workforces that are distributed globally, including Hofy, which was acquired by Firstbase, Workwize, and GroWrk. Dan Duggan joined the company as its cofounder after the pair met via the Y-Combinator co-founder match platform in 2024. Day 1: Displacing the Excel Sheet According to Osifo, the first version of Rayda was built to replace spreadsheets and give companies a single place to record their assets. “We were trying to displace Excel sheets in the place of asset management for growing businesses,” he said. Rayda’s first customers, Osifo noted, exposed a blind spot. He had assumed that businesses would arrive on the platform with spreadsheets ready to import into the platform. Instead, he found that some of these businesses had no asset register. “We saw that a lot of the small and medium-sized businesses had a lot of assets, but most of them didn’t have an Excel sheet in the first place,” he said. Instead of a regular software onboarding process, Rayda found itself sending staff to physically go to customer offices and audit what they owned before onboarding them onto the platform. Before he launched the product, Osifo noted that he spoke to other founders to validate the answers and received responses that were consistent enough to convince him there was a real business there. He noted that some of those conversations turned to early investments. He noted that he deliberately kept Rayda’s first raise small, targeting $250,000. The round eventually closed at more than $300,000 before the startup’s public launch in January 2023. “It made sense for our business, plus raising too much capital early on affects how a startup approaches problems: most problems then seem like a problem to spend money on,” Osifo said. “I think when you raise in the life of a business is as important as how much you raise.” Day 500: The pivot By the end of 2023, Rayda had transformed beyond the platform Osifo initially launched. He explained that customer feedback kept pulling the startup in new directions. Rayda Core remained its central asset management platform, but it was joined by Bidda, which helped companies auction unwanted office equipment; Rayda Track, which monitored devices; and Rayda Remote, a platform that let businesses procure and provide laptops for employees. “We ended up having almost like three or four other product expansions that we explored,” Osifo said. “Our pitch then was that you could buy, manage, track, and dispose of all types of fixed assets.” For a while, the strategy held up and reflected Osifo’s initial ambition to build an end-to-end platform. Then 2024 happened. In 2024, Nigeria’s currency experienced a sharp depreciation, falling by 129.23% and closing the year at ₦1,478.97 ($1.07) per US dollar. This was more than double the exchange rate from the previous year’s ₦645 ($0.47). This made businesses more cautious with spending. The consequences of macroeconomic issues forced Rayda to change the trajectory of its business. “We were always going to go global, but we were initially trying to own a significant market share in Africa before we thought about being global,” he said. “2024 forced us to change the business.” He explained that the company needed a new lever of growth and began experimenting with international expansion by exploring which parts of its product suite would resonate beyond Africa. “We weren’t really sure what was going to finally stick,” Osifo said. The answer came from a company in the United Kingdom, which he did not disclose, that had employees in Nigeria. Osifo explained that the company presented with the same problem he had noticed while at 54gene: it was taking them too long to get a laptop to a new hire in Nigeria. He explained that Rayda’s Remote platform solved the problem for the company. He explained that development on the other products paused upon seeing the volume that one customer generated from using its Remote platform. In the three
Read More
Klump brings multi-bank instalment payments to Jumia’s online checkout
Klump, a Nigerian buy now, pay later (BNPL) startup, has partnered with Jumia to bring instalment payments to one of Africa’s largest e-commerce marketplaces, expanding its strategy of embedding consumer credit directly into online checkouts. The partnership places Klump inside Jumia’s checkout, allowing shoppers to compare financing offers from multiple banks without leaving the platform. Rather than lending itself, Klump provides the technology that connects borrowers with partner lenders, which assess applicants, fund the loans, and assume the credit risk. Embedding its technology inside the e-commerce giant gives the startup access to high-intent shoppers while reinforcing its strategy of becoming the infrastructure layer for consumer credit rather than a lender itself. “When we started Klump, our mission was simple: give Nigerians access to affordable credit, wherever they shop,” Klump co-founder and CEO Celestine Omin wrote in a Friday LinkedIn post. “Today, we’re excited to partner with Jumia, bringing instalment payments to one of Africa’s biggest marketplaces. Klump now powers instalment payments for two of Nigeria’s largest e-commerce platforms.” The integration also expands Jumia’s BNLP offering, which previously included partnerships with CredPal and Easybuy, making Klump the latest fintech to embed consumer credit directly into the online shopping experience. Klump does not lend or hold credit risk, which sits with partner banks. Its strategy is that owning checkout placement across Nigeria’s biggest marketplaces matters more than owning a loan book, even if it means ceding control over approval rates and terms to lenders it does not underwrite. How it works The integration moves the loan application into the checkout, so customers no longer need to seek financing separately before making a purchase. According to Klump, shoppers choose eligible products before selecting “Pay with Klump” at checkout. They then select a financing provider, complete a credit assessment using their banking and identification details, and, if approved, pay an initial deposit of 20%-30% of the purchase price. The remaining balance is repaid in instalments. Shoppers must be at least 21 years old, have an active Nigerian bank account with regular salary or business income, and provide a valid government-issued ID. Depending on the lender, additional verification, including facial recognition or one-time password authentication, may also be required. Loan terms vary by lender Rather than offering a single financing product, Klump presents customers with multiple lenders that compete on loan size, deposits, repayment periods, and pricing. Customers can currently choose financing from First Bank, Renmoney, Credit Direct and Wema Bank, each offering different loan limits, deposit requirements and pricing, according to Klump. Customers can currently choose financing from First Bank, Renmoney, Credit Direct, and Wema Bank, according to Klump. Loan sizes range from ₦1,000 ($0.73) to ₦2.6 million ($1,898), with repayment periods of six to twelve months. Klump says it does not charge interest itself; each lender determines the loan pricing. Customers whose banks are not supported can still apply through lenders such as Renmoney and Credit Direct, which do not require applicants to hold accounts with them. Approved loans are not automatically cancelled if a customer returns an item. Klump says shoppers can instead use the value to buy another product at the same price or top up their Klump Wallet to purchase a more expensive item, rather than receive a cash refund. True scale demands moving beyond surface-level integrations to robust execution. We’ve filtered the noise out of Moonshot 2026, optimising the conference strictly for high-calibre connections between startup founders, global financial operators, enterprise leaders and individuals rewiring Africa’s technical frameworks. Get 20% off Early Bird tickets for a limited time.
Read More
One UI 9 features Samsung users need to know
Table of contents When is the One UI 9 release date? One UI 9 features you need to know Which phones are getting One UI 9 Samsung is rolling out One UI 9, its new software update built on Android 17. If you own a Galaxy phone or tablet, this update brings new tools for your camera, your lock screen, your security settings, and more. Here is what you need to know about the release date, the new features, and when your device will get the update. When is the One UI 9 release date? Samsung has already started testing One UI 9. The beta launched on May 12, 2026, for the Galaxy S26 series, and it is currently available in Germany, India, South Korea, Poland, the UK, and the US. The full version of One UI 9 debuts on July 22, 2026, at Samsung’s Unpacked event in London. Samsung will preinstall it on the new Galaxy Z Fold 8, Fold 8 Ultra, and Z Flip 8. If you own an older Galaxy phone, you will wait a bit longer. Reports point to a stable rollout for the Galaxy S26 and other eligible devices around September 2026. Samsung has not confirmed this date yet, so treat it as an estimate rather than a fixed schedule. If you live in South Africa or elsewhere in Africa, expect the update to arrive several weeks after it arrives in Korea, Europe, and the US. Samsung rolls out updates in waves, and African markets are usually not first in line. One UI 9 features you need to know One UI 9 does not change how your phone looks. It improves the tools you already use every day. Here are the features Samsung has confirmed so far, plus a few that leaked ahead of the official launch. A Quick Panel you can customise: Your Quick Panel now gives you more control. You can adjust brightness, sound, and your media player on your own instead of all together. You can also resize your media player into a small square without needing extra apps. A New Tape Tool in Samsung Notes: Samsung Notes now has a tape tool. You can cover part of a note the way you would with real tape, then peel it back when you need to see what is underneath. This works well if you are studying or hiding sensitive details. You also get new pen styles and a tool that automatically straightens your lines. Better multitasking with Samsung DeX: You can now move app windows between virtual desktops. You can also preview your desktops from the Recents screen and switch between them with one tap. New tools in Game Booster: While you play, you can change your screen resolution, pick your screenshot format, and monitor your FPS, CPU, and GPU. You do this without leaving your game. A Blue Dot for Location Tracking: Android 17 adds a blue dot that shows up when an app uses your location. Tap the dot in your Quick Panel to see which app is tracking you. You also get a clearer choice between sharing your approximate location and your precise location. Stronger protection against risky apps: Samsung is adding more safeguards to protect against harmful apps. Your phone can now warn you, block installation, or recommend you delete an app that looks suspicious. You also get a new menu that lists every app you sideloaded, so you can easily review or remove them. Text Spotlight for easier reading: Text Spotlight lets you tap on text to enlarge it in a floating window. You can change the font size and colours to make reading easier. One TalkBack for accessibility: Samsung merged its TalkBack feature with Google’s version. This means you get updates straight from the Play Store instead of waiting on Samsung. You also get a new Select to Speak tool that reads text and images out loud. A Call Log that shows every app: Your call log now shows calls from apps like WhatsApp and Google Meet, not just your regular calls. Samsung is also adding on-call context, so you can see recent messages or important dates about a contact while you are talking to them. Gemini Intelligence, coming later: Samsung’s biggest AI feature, called Gemini Intelligence, will not appear in the beta. Samsung confirmed it is coming with the new foldable phones. This AI can complete tasks for you, like booking something or drafting an email. It may only work on flagship devices, since it needs strong hardware to run. A few of these features are rumoured and could still change before Samsung finalises the update. Which phones are getting One UI 9 Samsung confirmed a wide list of eligible devices. Your update time depends on which series you own and where you live. Here is the full breakdown. Samsung is not bringing One UI 9 to the Galaxy S22 series or the Galaxy S21 FE. These phones have already received their final major update, One UI 8.5. Keep checking your software update settings as your wave approaches, since Samsung ties the exact date to your carrier and region. True scale demands moving beyond surface-level integrations to robust execution. We’ve filtered the noise out of Moonshot 2026, optimising the conference strictly for high-calibre connections between startup founders, global financial operators, enterprise leaders and individuals rewiring Africa’s technical frameworks.Get 20% off Early Bird tickets for a limited time.
Read More
Accrue targets African businesses with stablecoin-powered cross-border banking platform
Accrue, an agent-led stablecoin fintech that operates in several African markets, has launched a banking platform for small and medium-sized businesses, in a move to capture rising demand for faster, cheaper cross-border business payments. Accrue Business allows businesses to hold, send, and receive stablecoins, collect international payments, and pay suppliers across borders using the same infrastructure that powers Cashramp, the company’s consumer remittance product. Businesses can create a stablecoin wallet, virtual US dollar and euro accounts, invoicing tools, virtual cards, and stablecoin-based payroll, allowing them to receive international payments, pay suppliers across Africa, Europe, and the United States, manage employee spending, and pay staff directly into on-chain wallets from a single platform, according to Clinton Mbah, co-founder and chief executive officer of Accrue. “We have seen a lot of success with individuals who haven’t had a reliable, affordable, and fast mobile money way for them to send money across Africa,” Mbah told TechCabal. “Businesses started asking for the same thing.” The launch comes as more African fintechs adopt stablecoin rails to solve cross-border payment bottlenecks for businesses. Africa’s cross-border payments market processed about $329 billion in 2025 and could grow to reach $1 trillion by 2035, according to a report by venture capital firm Oui Capital. As the market expands, fintechs such as Accrue are positioning stablecoins to route a significant chunk of those transaction flows. The startup joins other fintechs such as Grey, Flutterwave, and Raenest, which offer stablecoin-powered payment products that enable businesses to collect international payments, hold dollar balances, and pay suppliers across multiple markets, reflecting a broader shift in how fintechs are approaching cross-border commerce. However, Mbah said Accrue is taking a different approach. It is building on an agent network that already moves money across 15 African countries, using stablecoins as the settlement infrastructure, while local agents provide liquidity on both sides of a transaction. “If I’m in Ghana and want to send money to Nigeria, I’m matched with a Ghanaian agent who converts my Cedis into stablecoins,” Mbah said. “When I send the payment, another agent in Nigeria settles Naira to the recipient in exchange for those stablecoins.” Accrue does not provide the counterparty liquidity itself; its agents directly do, bypassing traditional payment intermediaries and banking partners. That model allows the startup to charge businesses about 1% for transactions, while US dollar payments incur about 0.5% or higher, depending on customer volume, according to Mbah. He noted that the startup retains about 85% of the processed margins, while the remaining 15% goes to the peer-to-peer (P2P) agents processing the transactions. Mbah added that removing banks and payment processors from the transaction enables Accrue to cap Cashramp fees at $2 regardless of transaction size. To reduce counterparty risk, the fintech manages a closed network of vetted agents with established track records on its platform. It matches each payment request to an approved agent, who supplies the liquidity needed to settle the transaction, according to Mbah. He said agents continue receiving transaction flows only if they maintain a strong settlement record. Much of that network grows through referrals rather than formal recruitment. Mbah said Accrue’s agents—typically people aged between 18 and 30 without formal employment—earn upwards of $150 monthly, depending on how much starting liquidity they bring to the platform. Existing agents, many of whom have built long transaction histories on the platform, frequently recruit family members and friends into the network. Mbah said those referrals make it easier for Accrue to vet new agents as it expands into new markets while keeping customer acquisition costs low. “They [agents] have no incentive to be bad actors because for them this is a serious livelihood,” Mbah said. “They know that if they continue processing transactions successfully, they continue getting orders, and continued orders mean they earn more.” Accrue also extends stablecoin credit lines to agents with a strong track record, allowing them to process larger transactions without holding the full liquidity upfront, Mbah said. “We look at agents who have been on the platform for a long time, completed thousands of transactions, and processed several hundred thousand dollars in volume,” Mbah said. “We collect data on their activity and utilisation, and that determines how much credit we extend to them.” The credit is repaid automatically as those agents process new customer transactions, he added. The model has fuelled the startup’s growth. Cashramp’s transaction volume nearly quadrupled between 2023 and 2024 before doubling again between 2024 and 2025, according to Mbah. While he declined to disclose exact payment volume, he noted that Accrue has processed between $70 million and $100 million in total transaction volume across its platform since launch, including stablecoin on-chain deposits, withdrawals, and foreign currency inflows and outflows. Cashramp accounts for between $30 million and $50 million of that volume. The startup is now extending that infrastructure to businesses. Accrue Business currently processes between $600,000 and $700,000 in monthly transaction volume across customers using its application programming interface (API), web platform, and recently launched mobile app, according to Mbah. The web platform has been live for about two months, while the API has supported business customers since 2024. The startup targets sole proprietors and small businesses that move money across borders regularly and remain underserved by larger financial institutions. Mbah said transactions currently top out between $150,000 and $250,000, a ceiling that reflects the businesses Accrue serves today. “We have merchants who serve customers across multiple African markets, and our ever-expanding Cashramp Agent Network perfectly caters to this challenge by providing agents on the ground who supply local currency liquidity,” Nureni Imam, Accrue’s head of business, told TechCabal. “Some of our merchants previously used traditional payment partners, but were limited by the number of supported currencies and the associated fees.” Accrue’s expansion strategy follows the same playbook that built its consumer business. The startup now has agents across more than 15 African countries, up from about seven before raising a $1.58 million seed round in January 2025. Mbah said Accrue’s deepest liquidity pools are
Read More
This Nigerian microfinance bank’s slow-lending strategy is delivering fast results
Speed is a competitive advantage in Nigeria’s $2.1 billion digital lending market. Fintech lenders promise approvals in minutes, using automation to move borrowers from application to disbursement with as little friction as possible. However, Nombank, the microfinance banking subsidiary of Nigerian fintech Nomba, has taken a different approach after securing its licence in December 2024. Rather than approving loans within minutes, the lender said it deliberately takes between 24 and 48 hours to make a credit decision. Payment data powers much of its underwriting, but account officers still verify merchants before any loans are approved, sometimes through physical visits. That slower process, the bank says, has helped it keep its non-performing loan ratio below 1%, well below the banking industry’s average of 8%. “We want to give loans that will truly come back,” Seun Osunkeye, Nombank’s managing director, told TechCabal in an interview. “Most importantly, we are interested in businesses growing, merchants growing, and everyone being happy.” In April, Nomba and Globus Bank announced that their 18-month lending partnership had disbursed ₦21.3 billion ($15.45 million) to Nigerian businesses while keeping non-performing loans below 1%. Non-performing loans are loans that borrowers have stopped repaying, making them a key measure of a lender’s credit quality. A ratio below 1% means fewer than one in every 100 loans on Nomba’s books has fallen into default, suggesting the bank has kept bad debts unusually low even as it expanded lending. This is well below industry standards, where eight in every 100 loans fall into default. In its January economic report, the Central Bank of Nigeria noted that the industry’s non-performing loan ratio rose to 8.03% in January 2026, above the prudential threshold of 5%. Lending against cash flow Nombank’s lending model begins long before a customer applies for credit. Because it lends exclusively to merchants already using Nomba’s payment infrastructure, the bank has months of transaction data before it makes a lending decision. Osunkeye said merchants are typically expected to have processed payments on the platform for at least three months, allowing Nombank to observe how money flows through the business before extending credit. The bank notes that it analyses daily, weekly, and monthly transaction patterns flowing through merchants’ accounts, studying how consistently businesses generate revenue, whether sales fluctuate seasonally, and how stable their cash flow has been over time. Those patterns determine not only whether a merchant qualifies for credit but also how much the bank will be willing to lend. Osunkeye explained that the facilities are generally capped at about 20% of a merchant’s monthly inflows, and most are structured as seven-day, two-week, or one-month working capital loans designed to help businesses replenish inventory or meet short-term operating needs. Most of the loans disbursed go to retailers. “It is mostly in the retail segments. That is where we see this kind of need. If you are looking at the big guys, like big sectors, those are probably looking for expansion, but these other people, they just need to get one or two things quickly,” Osunkeye said. The bank currently lends between ₦50,000 ($36.26) and ₦1.5 million ($1,087.69) per customer, with an upper lending limit of ₦2 million ($1,450.25). Is your business eligible? Adjust your average monthly payment inflows to see your estimated working capital loan limit. Average Monthly Inflow ₦ Estimated Loan Limit ₦ 200,000 *Capped at Nombank’s maximum limit of ₦2,000,000. System Insight: Nombank strictly caps lending facilities at about 20% of a merchant’s monthly inflows. By enforcing this cap and utilizing account officers for 24- to 48-hour physical verifications, they have maintained a non-performing loan ratio of below 1%—substantially lower than the 8.03% banking industry average. Source: TC reporting, CBN Nombank’s strategy relies heavily on Nomba’s broader business model. As more merchants process payments through its terminals, the company gathers richer transaction histories, making it easier to assess risk and extend credit. Daily payment volume across Nomba’s platform has grown from about ₦7 billion ($5.08 million) in May 2025 to roughly ₦250 billion ($181.28 million) by May 2026, according to company figures shared with TechCabal. Why Nombank still knocks on doors Despite relying heavily on payment data, Nombank has taken a hybrid approach to its lending process, combining automation with human assessment. Before many loans are approved, account officers confirm that merchants are still operating, maintain regular contact with customers, and, where necessary, visit businesses physically before disbursement, according to Osunkeye. The MFB previously experimented with near-instant lending but concluded that transaction data alone was insufficient to assess repayment risk. “We tried ensuring customers could get loans in minutes,” Osunkeye said. “But you might assume you still have a relationship with a merchant even though the account officer hasn’t visited that business for a long time.” For Nombank, payment data identifies potential borrowers. Character profiling helps with the final lending decision. “So, for us, we prioritise 24 to 48 hours,” he said. “The account officer checks the merchant before we disburse the loan. Everyone is selling automated, quick loans, but that relationship is very important. Character is very important.” The process is slower than that of many fintech lenders, but Nombank believes the additional checks improve repayment rates and reduce credit losses. According to Gbemi Adelekan, president of the Money Lenders Association, digital lenders face severe structural and operational risks. “Such risk, including the exceptionally high non-performing loans, coupled with the high cost of technology in providing loans in minutes, allows money lenders to charge a little premium on their interest rates,” he told TechCabal. Scaling without losing discipline Nombank currently operates under a tier-1 microfinance banking licence, according to Osunkeye, limiting its operations largely to urban markets, and it must have a minimum capital requirement of ₦200 million ($145,025). The MFB is currently operating within the Ifo local government of Ogun State, a state that borders Lagos in southwestern Nigeria. The bank says it has disbursed about ₦500 million ($362,563) in loans so far this year after lending roughly ₦100 million ($72,512) throughout last year. It expects to
Read MoreAs algorithms shape financial inclusion, accountability must keep pace
While public attention remains fixed on artificial intelligence (AI) tools that write, generate, and automate tasks, a more consequential transformation is quietly unfolding across Nigeria’s digital economy: algorithms are beginning to make decisions that affect people’s financial lives. AI is increasingly being used to make decisions rather than merely assist humans. In Nigeria’s fast-growing fintech sector, algorithms are beginning to influence who gets access to credit, how customers are assessed, how fraud is detected, and how complaints are resolved. As these systems become more sophisticated, a critical question emerges: who is accountable when an algorithm makes the wrong decision? This question is particularly important in Nigeria because financial services are becoming increasingly digital. According to EFInA’s 2023 Access to Financial Services Survey, formal financial inclusion in Nigeria rose to 64% in 2023, up from 54% in 2020, while overall financial inclusion reached 74%. More Nigerians now access financial services through digital channels, meaning decisions once made by human loan officers are increasingly being delegated to automated systems. AI-powered systems process applications faster, reduce operational costs, detect fraud more effectively, and expand access to financial services. For a country still working to deepen financial inclusion, these innovations offer enormous potential. However, efficiency is not the same as accountability. Unlike traditional credit decisions, where a loan officer can explain the factors that informed an outcome, AI-driven decisions are often based on proprietary models controlled entirely by the financial institution. Customers typically have no visibility into the metrics being used to assess them, no meaningful opportunity to challenge those metrics, and little understanding of how a decision was reached. This creates a significant information imbalance between service providers and consumers, undermining fundamental consumer protection principles such as transparency, fairness, and the right to seek redress. Consider a digital lending platform that uses artificial intelligence to determine creditworthiness. An applicant is denied a loan despite having a stable source of income and a history of responsible financial behavior. When the customer seeks an explanation, the company may be unable or unwilling to provide one beyond a generic statement that the application did not meet the platform’s risk criteria. Yet the algorithm, the variables it relies on, and the weight assigned to those variables remain entirely within the control of the institution. From the consumer’s perspective, this is more than a technology problem. It is an accountability problem. Who bears responsibility in such a scenario? The company may argue that the system operates automatically. The software developer may insist that it merely provided the technology. The data provider may claim responsibility ends with supplying information. Yet from the consumer’s perspective, someone must be accountable. What makes the Nigerian situation particularly significant is the speed at which technology adoption often outpaces regulation. According to the Central Bank of Nigeria’s Fintech Report, close to 11 billion transactions were processed through the NIBSS Instant Payment platform in 2024, more than double the approximately five billion recorded in 2022. The report also notes that AI is already being deployed across the fintech ecosystem. Nigeria’s Data Protection Act already provides an important foundation for addressing some of these concerns. Because AI systems rely heavily on personal and behavioral data, organisations must comply with obligations relating to lawful processing, transparency, and accountability. However, data protection alone cannot answer the broader governance question. That scale makes AI accountability a practical concern, not a theoretical one. A flawed employee may affect dozens of customers; a flawed algorithm can affect hundreds of thousands. The real challenge is ensuring that organisations remain responsible for decisions made through automated systems. As businesses adopt advanced AI tools, there may be a temptation to treat algorithmic outcomes as neutral or objective. That would be a mistake. AI systems are designed, trained, deployed, and monitored by human organisations. Responsibility cannot disappear simply because technology is involved. Nigeria has an opportunity to act before a crisis forces action. Rather than waiting for widespread disputes over automated decisions, policymakers now develop a clear framework for AI accountability in sectors where algorithms affect people’s access to services, opportunities, and financial resources. The principle should be straightforward: automation should never eliminate accountability. Companies should remain responsible for decisions made through systems they deploy. Consumers should have access to meaningful explanations where automated decisions significantly affect them. Regulators should be able to identify who bears responsibility when harm occurs. This does not require an immediate, sweeping AI law. It requires clear rules that preserve innovation while ensuring that businesses cannot hide behind algorithms when things go wrong. The goal should be responsible deployment, not regulatory paralysis. Nigeria has often had to regulate emerging technologies after problems have become widespread. AI offers a rare opportunity to take a more proactive approach. ___ Omoruyi “Uyilaw” Edoigiawerie is a startup lawyer, policy advisor, and Founder of EandC Legal. He advises startups, investors, and institutions on technology, innovation, and regulatory strategy. His work focuses on technology governance, digital transformation, financial inclusion, and emerging technologies, with a particular interest in the policies shaping Africa’s digital and economic future.
Read More
Freddie Omany: The French teacher who ended up moving millions of payments
“Be anything.” It sounds like the sort of advice printed on a graduation card. Freddie Omany’s older brother meant it literally. Long before he found himself helping move millions of mobile money transactions across Africa every day, Omany, PawaPay Kenya country director, taught French. He translated documents. He worked in government communications. He chased unpaid invoices for Booking.com, a digital travel platform, across Francophone Africa. None of those jobs looked remotely connected at the time. Looking back, he insists they were all preparing him for the same thing. We meet on the sidelines of Road to Moonshot in Nairobi, where PawaPay is one of the event partners. He is difficult to categorise. Payments executives usually talk in numbers: uptime, transaction values, fraud rates, APIs. Omany reaches instead for stories. Ask about financial inclusion, and he talks about a Bolt driver waiting to pay school fees. Ask about reliability, and he remembers a GiveDirectly—a non-profit that helps donors send money—recipient whose cash transfer could not afford to fail. Ask about leadership, and he quotes a mentor and not a management book. The longer we talk, the clearer it becomes that he does not think of payments as moving money. He thinks of them as moving trust. That may explain why the most interesting thing about Omany is not that he helps oversee one of Africa’s busiest payment networks. It’s that, after all these years and all these careers, he still approaches business like the curious boy from Molo, a small agricultural town 250km west of Nairobi, who kept asking why. Our conversation wanders far beyond payments. We talk about growing up in Molo, teaching French, why Africa’s payments revolution is still unfinished, and why he believes the best infrastructure is the kind nobody notices. This interview has been edited for length and clarity. Let’s begin before fintech. What sort of child were you? What fascinated you growing up, and what kind of family shaped you? Molo. That’s where it starts. A small farming town in the Rift Valley, a couple of hundred kilometres from Nairobi, and famously cold. People don’t always believe me when I tell them there are places in Kenya where you can see your breath in July. It was quiet, and Nairobi felt very far away. My family was a normal Kenyan family, and I mean that in the best sense. Nothing dramatic, nothing that would make a good founder origin story. What they gave me was room to be curious. I was the child who asked why about everything and took things apart to see how they worked, usually without permission and often without successfully putting them back together. That instinct never left. Everything I’ve done since, and I’ve done a lot of different things, comes back to the same reflex: here is a problem, let’s figure it out. Was there a moment when you realised business, or solving commercial problems, was more exciting than following a conventional career? It crept up on me rather than arriving as one big moment. Believe it or not, I’ve been a teacher. I taught French and worked as a translator. I’ve done communications for a government fund. Then I spent time at Booking.com managing finance across Francophone Africa, calling hotel owners from Dakar to Antananarivo about unpaid invoices. Somewhere in those calls, I noticed I wasn’t tired at the end of the day. Every invoice was a puzzle. Sometimes it was a cash-flow problem, sometimes a currency problem, sometimes purely a trust problem. Untangling it felt a lot like teaching, actually. You meet people where they are and move them somewhere better. Once I saw that, the idea of a conventional career stopped being interesting. I never really followed conventional career advice anyway. What advice did you ignore when you were younger, and would you make the same decision today? The advice is to pick one lane and stay in it. Everyone tells you to specialise early. I ignored that completely, and my brother deserves the credit. He told me to be a master in B.A. Not Business Administration. Be Anything. So I have been a teacher, a communications person, a finance manager, a customer success lead, a strategic operations manager, a partnerships manager, and now a country director running Kenya and South Sudan for PawaPay. Every one of those looked like a detour at the time and turned out to be preparation. Would I make the same decision today? Faster. Looking back, which failure taught you more than any promotion ever did? There was a time we spotted a problem on a partner’s network before their own team did. Our monitoring picked up a pattern that usually comes before a bigger degradation, so we made the call, rerouted traffic, and flagged it to them. Technically, we were completely right. Two hours later, it played out exactly as predicted. But the way we delivered it made the partner feel accused rather than supported. The awkwardness of that conversation took longer to repair than the incident itself. Being right was not enough. That failure crystallised something I now hold as a rule: people care about how you made them feel more than almost anything else, even in business. A merchant remembers the support team that solved their issue in record time. A partner remembers that you reached out when their CEO was unwell. Nobody frames the invoice. We are in the business of being human. No promotion teaches you that. Failure does. Omany presents during the Road to Moonshot event in Nairobi on July 2. Image source: TechCabal You’ve worked through different phases of Africa’s payments evolution. What has surprised you most about how money actually moves across the continent? Not what reports say, but what you’ve personally witnessed. How impatient this continent is, in the best possible way. Africans love speed. Fast payments, reliable payments. A trader in Kinshasa and a farmer in Addis Ababa have the same expectation: money should move now, and it should arrive complete. That
Read More
South Africa-founded startup launches AI model with 10 million-token memory
Refiant AI, a South Africa-founded startup that uses algorithms to compress artificial intelligence (AI) models, has launched Protea, a suite of large language models (LLMs) that the company says is capable of processing up to 10 million tokens in a single prompt. The models can process and retain significantly more information at once before generating a response, according to the company. Protea comes in three versions with context windows of one million, five million, and 10 million tokens and can be accessed for free through Refiant’s platform without a waitlist or approval process. The launch comes three months after Refiant AI raised a $5 million seed round led by VoLo Earth Ventures to expand its AI optimisation platform and deepen research partnerships. Protea marks Refiant’s first major product launch since its fundraising. “Customers don’t need more waitlists,” Mathew Haswell, Defiant AI’s cofounder, said. “They need models they can test, break, and build with. Protea is live, and we want people to use it from day one.” Founded in 2025 by Haswell, Viroshan Naicker, and Siddharth Gutta, Refiant AI is building machine learning systems to reduce compute costs and improve AI model efficiency. According to the company, at its maximum capacity, Protea can hold roughly 7.5 million words, allowing users to analyse large volumes of information without splitting it into smaller chunks. It added that such a level of context could allow legal teams to review hundreds of contracts in a single pass, insurers to analyse years of claims data, and engineering teams to process entire software codebases. The launch comes as AI companies increasingly compete to expand the amount of information their models can process at once. Anthropic’s Claude supports context windows of up to 500,000 tokens on certain Enterprise offerings, while Google’s Gemini offers up to one million tokens on its higher-tier plans. As AI developers move to improve performance by expanding context windows, Refiant is betting that a larger AI memory will give enterprises an edge when analysing large datasets. “Long-context AI has been talked about for over a year now, but hasn’t really been commercially available,” said Naicker, CEO and co-founder of Refiant. The company said it has already demonstrated an internal prototype capable of processing up to 100 million tokens and is exploring how to benchmark and bring the technology into production. It added that Protea is the first of three planned product releases. True scale demands moving beyond surface-level integrations to robust execution. We’ve filtered the noise out of Moonshot 2026, optimising the conference strictly for high-calibre connections between startup founders, global financial operators, enterprise leaders and individuals rewiring Africa’s technical frameworks. Get 20% off Early Bird tickets for a limited time.
Read More