Identity verification is a major problem for financial services companies in Nigeria, and we can overcome that challenge by using Amazon Rekognition. That gives us a competitive edge as a startup.
Wale Akanbi CTO, Aella Credit
  • Company: Aella Credit
  • Industry: Financial Services
  • Country: Nigeria
  • Employees: 50
  • Aella Credit is a financial services technology startup that provides easy access to credit to the world’s underbanked. The company is based in Lagos, Nigeria; San Francisco, California; and Manila, the Philippines. Aella Credit provides machine learning–driven risk assessment in both a B2B integration with employers/cooperatives and a B2C model to determine applicant eligibility for loans.
  • Improves facial recognition accuracy by 40%
  • Increases availability of loan processing software
  • Grows from 5,000 to 200,000 customers in several months

Aella Credit is a financial technology company dedicated to providing instant loans to Africans through a mobile loan application platform. As it sought to innovate and grow faster, Aella Credit found itself limited by its existing technology environment. “We launched the company in the cloud, but the technology we used was holding us back,” says Wale Akanbi, chief technology officer for Aella Credit. “It didn’t offer us the innovative features we needed.” For example, the company wanted a better way to validate employee IDs and government-issued IDs in real time, without human intervention. “In Nigeria, there is no API to use to verify government IDs in real time”, Akanbi says. “This made the verification process tedious for us, and we had multiple overlapping user profiles and duplicate datasets.”

To address the problem, Aella Credit decided to search for a new cloud solution that was more reliable and could support the organization's web-based loan-processing software. As it embarked on the search, the company needed to ensure it found technology that was more reliable and support the organization’s web-based loan-processing software. “We can’t afford to have any downtime, because we’re dealing with our customers’ money,” Akanbi states.

Although it considered several cloud technology providers, Aella Credit chose Amazon Web Services (AWS) as its new provider. “AWS met our needs perfectly in terms of innovative features and scalability,” remarks Akanbi. Aella Credit runs its online loan- processing solution on Amazon Elastic Compute Cloud (Amazon EC2) instances and takes advantage of Amazon Relational Database Service for MySQL (Amazon RDS for MySQL) to support database instances running several versions of MySQL software. “Support for MySQL was one of the key reasons we decided on AWS,” confirms Akanbi.

Aella Credit is also using Amazon Rekognition, a deep learning technology solution that enables the addition of image and video analysis to applications. Aella Credit uses Amazon Rekognition for facial detection for new customers. When customers upload a profile image to the mobile app, the image is sent to Amazon Rekognition and saved in Amazon Simple Storage Service (Amazon S3). The customer’s facial expression is analysed and saved in Amazon RDS.

In addition, when customers update their profile pictures, the new profile image is sent to Amazon Rekognition for comparison. If there is a match, the customer can proceed with updating or repeat the process if there is no match. For customers applying for large loans and employees of Aella Credit’s partner companies, Amazon Rekognition compares government-issued IDs with a base sample. When these customers or employees upload their IDs, they are compared with sample IDs. If the images match profile images, the customers or employees are granted access to credit.

Using Amazon Rekognition, Aella Credit is innovating by taking advantage of cutting-edge deep-learning and AI technologies, which has helped the organization address its identity verification challenges. For example, with Amazon Rekognition, Aella Credit has improved the accuracy of face verification by more than 40 percent. “This technology is very good at identifying customers based on their eye color or facial expressions, and we are using data analysis to recognize patterns that can help us better protect customers and their data,” Akanbi says. “Identity verification is a major problem for financial services companies in Nigeria, and we can overcome that challenge by using Amazon Rekognition. That gives us a competitive edge as a startup.” Due to its improved identity verification, Aella Credit can grow faster because customers are more confident in the bank’s security capabilities.

Aella Credit is also ensuring high availability for its loan-processing software and customer databases. “We have had zero instances of system downtime since we moved to AWS and started using MySQL on Amazon RDS,” states Akanbi. “AWS is incredibly reliable, so that gives us more time to focus on innovation rather than managing infrastructure.”

The company has grown since being on AWS, expanding from Nigeria to other emerging markets. “We went from 5,000 customers to 200,000 customers in just a few months because AWS helps us scale very quickly,” says Akanbi. “And even if we scale to a million customers at some point, we won’t have to worry about managing our architecture ourselves. We can just focus on developing new features to further differentiate our business.”

To learn more, visit: Machine Learning on AWS.