Maya, a Philippine fintech, is accelerating customer verification and onboarding by using the National ID system together with artificial intelligence, while keeping fraud safeguards intact.

When National ID details are incomplete, the app requests only the missing information and users can still opt for the traditional ID-upload path without restarting their application.

The AI, described as agentic, follows predefined workflow steps to support KYC, deduplication, information validation, checks, and exception routing to human reviewers, all under ongoing oversight.

Automation handles mechanical tasks, freeing analysts to focus on decisions requiring expertise, with governance and traceability preserved.

Onboarding turnaround time has fallen by about 84%, allowing staff to concentrate on exceptions, investigations, and cases needing human judgment.

Consumer onboarding time improved similarly by 84%, reducing customer wait times and enabling staff to address exceptions and investigations.

Screening noise remains a challenge, with high false-positive rates reported, prompting analysts to review many non-issues.

Generative AI can draft SAR/STR narratives from structured case data, provided all statements trace to underlying evidence, with analysts reviewing rather than drafting from scratch.

The broader vision is to redesign processes around trusted digital infrastructure to make financial services faster and more accessible while preserving human accountability and regulatory safeguards.

ML-based entity resolution reduces false positives by a wide margin and improves recall and F1 scores through fuzzy matching and contextual scoring.

Cross-border operations benefit from platforms linking to over 210 local registries, creating a consistent automated intake layer despite jurisdictional format differences.

AI is transforming KYC/AML compliance by automating repetitive data handling while preserving essential human oversight and decision-making.