TRI NGUYEN MICROSOFT INTERNAL
← Back to Overview
Compliance AI $18M+

Maybank Compliance AI Transformation

Multi-Phase AML & Name Screening Journey

$18.3M

Hub Pipeline

$8M+

Total Influenced

~30% reduction

Review Time

$70K+/yr

Productivity

The Story

The Maybank engagement was one of the most technically complex and strategically significant multi-phase journeys in FY26. Working alongside Lead Hub SE Sally Chin and MY account teams, I served as the core technical architect driving compliance AI innovation.

Phase 1 (Sep 2025): In-person half-day workshop on compliance operating model, Name Screening & KYC pain points. Anchored Name Screening as first build candidate.

Phase 2 (Jan 2026): Designed entity resolution patterns, semantic search vs. deterministic rules using Azure AI Search & vector embeddings. Co-authored the Maybank Compliance High-level Architecture.

Phase 3 (Feb 2026): Led deep technical sessions on vector embedding strategy vs. fuzzy match for regulatory workflows. Mapped analyst journey across fragmented systems. Prioritized AML Narrative Copilot design.

Featured in Innovation Hub Asia Impact & Outlook Report. Vera Siertsema called out 'strong collaboration in making this win happen'.

Key Workstreams

Name Screening Solution Envisioning — Entity resolution, vector embeddings

Transaction Monitoring Architecture — Mantas/ECM/STR flow mapping

AML Narrative Copilot Design — ML + GenAI hybrid architecture

Data Vectorization Deep Dive — Semantic similarity vs. fuzzy match

Compliance High-Level Architecture — Fabric + OneLake + GoldenGate