Kenyan startup PEMiG is a credit intelligence platform built specifically for African lenders, helping people with little or no formal credit history access loans.
Founded in 2022 by Felix Sifuna, Philip Amwata, Priya Maharaj and Lilian Kariba, PEMiG has developed a proprietary causal credit scoring engine (CCSE) system that uses causal AI, social capital data analytics, and behavioural credit scoring to help banks, MFIs, SACCOs, and fintech lenders assess borrower risk far beyond what traditional credit bureau data can reveal.
“The problem we solve is fundamental – over 80 per cent of adults in Sub-Saharan Africa are thin-file or no-file borrowers; people with little or no formal credit history,” Sifuna said. “Traditional credit scoring models were simply not built for this reality, which means lenders are forced to either take on poorly-understood risk and suffer rising non-performing loans, or exclude a massive segment of creditworthy borrowers and leave growth on the table.”
PEMiG’s CCSE system changes that equation. “Instead of asking “what does this person’s credit history say?”, we ask “who is this person really, and how likely are they to repay?”. We build dynamic, real-time borrower profiles that evolve over time, giving lenders a living picture of risk, not just a snapshot,” said Sifuna. The result is better underwriting decisions, reduced defaults, improved portfolio at risk (PAR) management, lower credit loss provisioning, and safe expansion into previously untapped borrower segments.
“When we looked at how lenders were operating, we saw institutions turning away creditworthy borrowers every single day; not because those borrowers were risky, but because the scoring tools couldn’t see them,” Sifuna said. “Meanwhile, the loans that were being approved were sometimes going to the wrong people, because traditional signals were being gamed or were simply insufficient. That asymmetry, exclusion on one side, invisible risk on the other, is the gap PEMiG was built to close.”
There are a handful of alternative credit scoring fintechs emerging across the continent, but he said most of them rely on mobile money transaction data alone, which still excludes large segments of the population and lacks the causal reasoning layer that makes PEMiG’s assessments genuinely predictive rather than correlative. “Our CCSE system is, to our knowledge, the only credit intelligence engine in the East African market deploying causal AI specifically for lender-facing risk decisioning,” Sifuna said.
Uptake has been “stronger than expected”. The startup already has five active partner lenders currently using its CCSE credit intelligence system, and two formal pilots running with additional institutions. Over 1,500 borrowers have been assessed through the platform, and US$250,000 in loan volume processed over 18 months.
“For a bootstrapped, early-stage platform selling into regulated financial institutions, which are notoriously slow to adopt new technology, we consider this meaningful early traction. Every lender that goes live with PEMiG has come through direct relationship-building and the strength of the product demonstrating real results. No shortcuts,” Sifuna said.“
We made a deliberate choice early on to prove the model first and grow from revenue, which has kept us disciplined, close to our customers, and deeply embedded in the real operational realities of African lending.
Kenya was the natural starting point for PEMiG, he said, as it has one of Africa’s most sophisticated and competitive lending ecosystems, spanning commercial banks, MFIs, SACCOs, and a thriving fintech lending sector. “If your product can earn the trust of Kenyan lenders, it can earn the trust of lenders anywhere on the continent,” he said.
“Our expansion roadmap targets East Africa, Uganda, Tanzania, and Rwanda, within the next 12 months. From there, the vision is pan-African: a continent-wide credit intelligence layer that works as well for a tier one bank in Lagos as it does for a community SACCO in Lusaka.”
PEMiG operates on a B2B SaaS model, charging lenders for access to its CCSE credit intelligence system on a subscription basis, with additional transaction-based fees tied to the volume of borrower assessments processed. “This aligns our revenue directly with lender usage and growth; when our partners grow their loan books, we grow with them,” said Sifuna.
“We are currently generating US$5,000 in monthly recurring revenue (MRR) – early-stage, but meaningful for a bootstrapped company selling complex technology to regulated institutions. Our focus right now is deepening value with existing partners and converting our active pilots into full contracts, which will drive the next step-change in revenue.”
He said the hardest thing about selling credit intelligence to African lenders isn’t the technology, but rather trust. “Banks and MFIs in Kenya have been burned before by technology vendors who over-promised and under-delivered, which means even a compelling product demo only gets you so far,” said Sifuna. “We spent months in conversations with institutions who loved what they saw but needed time and proof before they’d let us anywhere near their loan decisioning process. Patience and persistence are not optional in this business.”
Another major challenge is data, as building a causal AI model requires not just data, but the right data – clean, labelled, longitudinal data on social capital and borrower behaviour that simply doesn’t exist in structured form in most African markets. “We had to build significant data infrastructure from scratch, often working directly with lender partners to structure and clean historical loan performance data before we could even begin modelling. That groundwork took far longer and cost far more than we originally anticipated,” Sifuna. And though PEMiG has successfully bootstrapped, that hasn’t always been glamorous. “Building deep fintech infrastructure as a bootstrapped team in Nairobi without the runway that VC-backed competitors enjoy meant making hard calls constantly about where to spend time and money,” Sifuna said.
“There were months where the founding team went without pay to keep the product moving forward. That’s not a comfortable story to tell, but it’s an honest one, and it’s part of why we’re proud of what we’ve built.”He said what kept them going was simple. “Every time a lender partner told us that PEMiG had helped them approve a borrower they would have previously declined and that borrower repaid on time we knew we were building something real,” Sifuna said.