Lendsqr one of the leading global providers, focused on helping lenders build effective credit products, has officially launched in Tanzania.

As part of this entry, the company is introducing a new Artificial Intelligent-based voice and video analysis model, designed to support better loan assessments for borrowers who don’t have formal financial records.

Tanzania has a growing demand for credit, yet many people remain excluded because they don’t operate within the formal banking system.

From motorcycle taxi (boda-boda) riders and food vendors to Dukawalla owners, many Tanzanians work in cash and don’t have pay slips, bank statements, or credit histories.

Traditional credit scoring methods rarely work in these cases.

“With this launch, we want to help lenders answer one of the oldest questions in credit: will this borrower repay?” said the Chief Executive Officer of Lendsqr, Adedeji Olowe.

 “A food vendor (mama-lishe) or boda-boda rider in Mwanza might never have walked into a bank, but that doesn’t mean they’re not creditworthy. They just need a system that recognizes them.”

For years, lenders have relied on proxy indicators like bank statements, call records, or even GPS data to estimate a borrower’s capacity and character. But such data can be invasive, unreliable, or simply unavailable.

As Olowe explains, “The core questions are always the same: Can this person afford the loan? And will they pay it back? But the tools we’ve been using to answer these questions don’t work for everyone.”

In an ambitious attempt to address this credit gap, Lendsqr has initiated an experimental AI model, developed using Google Cloud technology, which evaluates how people speak and respond during loan applications.

It looks at tone, confidence, clarity, and other signals that can give lenders useful information, especially when conventional data points are missing.

This project is partly funded by the Nigerian government via the Ministry of Communications, Innovation and Digital Economy and Google.

The tool has already shown early success. In its first trial, the model delivered up to 76 percent accuracy in predicting repayment behavior.

Now in its second phase of testing, it is being made available to lenders in Tanzania, who often face challenges when assessing informal or cash-based borrowers.

“This is not about replacing credit officers or acting like a lie detector,” said Olowe. “It’s about giving lenders another way to make better decisions when formal data just isn’t there.”

Lendsqr acknowledges the ethical questions such a system presents, especially around privacy and bias. But, the mission remains clear.

Without new models, millions remain excluded from formal credit systems or are forced into exploitative loan arrangements.

“If we get this right,” said Olowe, “we could unlock meaningful, responsible credit access for East African adults who’ve been invisible to the financial system for too long.”

With this launch, Tanzanian financial institutions, whether fintechs, SACCOs, or traditional banks can use Lendsqr’s tools to offer credit through mobile apps (Android and iOS), web platforms, or direct integrations with their own systems. The platform also connects to third-party services like credit bureaus, payment processors, and KYC providers.

Lendsqr is a global Lending-as-a-Service (LaaS), that provides digital tools for lenders who want to build and manage credit products for individuals and small businesses.

With operations in Rwanda, Nigeria, the UK, the US, and the Caribbean, Lendsqr offers lenders everything from loan application software to decision-making tools and risk assessment models.

The platform supports different loan types, including personal and SME loans, and works for both online and offline borrowers.

It offers features like equity contributions, guarantor options, and manual loan processing for those without internet access.

A free trial is available for lenders looking to test the system before making long-term commitments.