Nigeria's banks move millions of transactions a day, and every SAR they file lands with regulators headquartered here. We build the compliance infrastructure to match, with the people who spent their careers writing those reports by hand.
In November 2023, Gift's team at Zenith Bank received 4,200 AML alerts in a single month, the highest ever, driven by new transaction-monitoring thresholds. His team of 12 investigators filed 38 SARs and dismissed the rest. The dismissal documentation alone took more time than writing the SARs.
He called Chidi, who had just left Google and was looking for a problem worth solving. “I need something,” Gift said, “that can look at an alert, pull the evidence, and either explain why it's fine or write me a SAR I can submit. Both need to be good enough to show a regulator.”
Yewande built the first prototype in three weeks. The skeptic agent was Chidi's idea.
CEO & Co-founder
9 years in compliance at Zenith Bank, last as Head of AML/CFT Operations. Left in 2023 after building an internal alert-triage tool that cut review time 40% but could not get it productized. CAMS-certified. Owns product vision and CBN/NFIU relationships.
“I've written thousands of SARs by hand. Every one follows the same structure, cites the same type of evidence. The only thing that changes is the specific facts. That is a solvable AI problem.”
CTO & Co-founder
BSc Computer Science, Obafemi Awolowo University. Senior engineer at Kuda Bank (fraud detection) and previously Interswitch. 6 years in fintech with a focus on graph-based fraud detection. Owns AI/ML architecture and multi-agent orchestration.
“The skeptic agent was our breakthrough. Until we built something that would try to disprove its own findings, the SAR quality wasn't good enough. Now it is.”
Head of AI Research & Co-founder
MSc Machine Learning, University of Edinburgh. 3 years at Google Zurich on abuse detection; returned to Abuja in 2022. Co-authored two papers on graph-based money-laundering detection. Owns agent architecture and adversarial validation.
“Financial crime AI fails when it optimizes for recall at the expense of precision. Every false SAR wastes a regulator’s time and erodes trust. We design for both.”
Compliance Specialist
5 years of NFIU reporting experience. Joined December 2024. Provides ground-truth review of generated SARs against the NFIU Filing Guidelines.
Founding insight: 4,200 alerts in a single month at Zenith Bank.
Team assembles; first prototype begins.
OmniVigil Technologies Ltd. incorporated in Abuja (FCT).
Pre-seed round (₦25M) closed in three weeks.
First pilot: tier-2 commercial bank, 500 alerts/month.
3 pilot customers, 28,000 alerts triaged.
340 SAR drafts accepted by pilot compliance officers.
CBN regulatory sandbox application; NFIU direct-filing API.
The SAR narrative exists to explain the evidence. We never generate a claim that is not linked to a specific, identifiable evidence item.
Our adversarial validation agent is not a feature. It is the reason our output is trusted by compliance officers who have been burned by automation before.
We work with the CBN and NFIU because our users file SARs with them. Regulatory trust is a product requirement, not a sales function.