Security24.04.2026

Two South African women developed software that protects R82 billion from fraud in a month

By Luis Monzon

Thalia Pillay and Carla Wilby met over a decade ago at the University of Cape Town, where they studied engineering before founding Cape Town anti-fraud startup Orca Fraud.

Like its marine hunter namesake, Orca’s real-time transaction monitoring uses a coordinated, “always-learning” approach to detect and stop fraud as soon as it occurs.

In March, the company raised R38.5 million in seed funding to expand its real-time transaction monitoring service, its primary product, and fraud intelligence across Africa and Latin America.

During their time as software engineers at Cape Town-based fintech Stitch, Pillay and Wilby saw that the rapid growth of financial technology across Africa was accompanied by growing fraud.

“What we saw was with all that growth, and more digital financial inclusion, we were also seeing more fraud,” Pillay said in a recent interview.

“The main problem we saw was that fraud in Africa was different. This is due to two reasons, the first being different payment methods, and then very different socio-economic climates.”

Pillay gave the example of Safaricom’s fintech darling M-Pesa, and how its enormous growth was not accompanied by a rapid growth in fraud compliance.

She has a background in robotics engineering and earned her degree cum laude. Before university, she received nine distinctions from St Mary’s Diocesan School for Girls in Durban, KwaZulu-Natal.

“We just started seeing a lot of real-time fraud on alternative payment methods,” she told MyBroadband.

“When we looked at different solutions on the market, we found that a lot of them didn’t fit the South African payment ecosystem.”

The pair started Orca Fraud in January 2024, with Pillay serving as the chief executive officer and Wilby as the chief technology officer.

Wilby has a bachelor’s degree in electrical and computer engineering (honours) and has a background in education technology, having founded e-learning platform Zomila in 2020.

Protecting companies in over 70 countries from fraudulent transactions

Wilby explained that the software uses “a combination of proprietary algorithms, heuristics, and machine learning” to determine the likelihood that a transaction is fraudulent.

“We are then able to return that to the provider, who is then able to make a decision on whether or not to proceed,” she said.

At the time of publication, the company protected over R82 billion in monthly transactions from fraud across more than 70 countries, working with banks, telcos, and payment providers.

Orca Fraud announced a partnership with Ozow, one of South Africa’s biggest payment providers by volume, late last year.

Ozow stated that it controlled over 70% of South Africa’s instant, third-party electronic payments as of November 2021.

“This partnership represents a shared commitment to building a safer, frictionless payments ecosystem where merchants and customers can transact with confidence,” it said in a LinkedIn post.

Orca’s latest funding round was led by tech growth fund Norrsken22, a returning investor, with OneDayYes, Enza Capital and CV VC Africa joining the round.

“Since our initial investment, Orca has evolved into critical infrastructure that enterprises increasingly rely on to manage fraud in high-velocity payment systems,” said Nivesh Pather, principal at Norrsken22.

“As digital payments accelerate and fraud becomes more organised and technology-driven, institutions need intelligence embedded directly into transaction flows to protect customers without slowing payments.”

The company now has a small team, with Pillay saying this was by design to keep it “tight, technical, and deeply embedded with clients.”

She said that as they scale across emerging markets and payment systems, their focus remained on ensuring safety keeps pace with scale.

Last month, Orca announced a partnership with African business-to-business-focused fintech Cauridor. The company operates cross-border payments between Africa and Canada and the United States.

Pillay said that the partnership allowed them to build models that can learn from the dynamically changing, context-specific payment channels and the fraud patterns that emerge from them.

“To build models that actually work in these environments, you need to learn from real transaction data where that fraud is happening,” she said.

Partners like Cauridor give Orca diverse, multi-corridor signals across African and American money flows, sharpening its interpretation of behaviour and protection of payments in emerging markets.