AEGI uses a deployed operating workflow to manage bounded evaluations from intake and readiness through ownership, review and handoff.
AEGI Operations v0.4.0AEGI Labs
Engineering governed risk-intelligence systems for credit and fraud workflows.
AEGI Labs builds AEGI Shield and AEGI Core. Shield works alongside existing risk systems to bring approved intelligence into governed review and bounded evaluation workflows; Core verifies declared evidence within scope.
Bring One Evaluation QuestionAEGI Labs capability
Product engineering backed by separate proof families.
Engineering, empirical evaluation and operating capability are linked to the evidence that directly exercises each one.
Controlled empirical evaluation under a frozen, declared benchmark scope.
RUN-BAF-003Mechanism and evidence-control behaviour within declared scope, with Core explained separately.
RUN-SCE-001NEXT EVIDENCE STAGE
Institution-specific value is assessed through a bounded customer engagement using the institution's declared workflow, baseline, approved evidence and review constraints.
BASEHo Chi Minh City, Vietnam
REGIONAL FOCUSSoutheast Asia
ENGAGEMENTBounded evaluations · Partnership and pilot discussionsDiscuss partnership or pilot
CAPITALOpen to strategic investment discussions with aligned investorsStrategic investment discussion
Founding team
Direct accountability across product, experience and systems engineering.
The people responsible for AEGI Labs remain visible alongside the product, systems and evidence they directly own.
Hoang Minh Truong
AI Product / Governance Architecture Lead — Vietnam
Product semantics, authority boundaries, Core–Shield architecture, evidence model, claim governance and strategic roadmap.
The Tinh Nguyen
AI Product Experience & Interactive Systems Lead — Vietnam
Human–AI review experience, evidence visualisation, interactive prototyping and programme/product communication.
Tri Minh Nguyen
AI Systems Architecture & Integration Lead — Vietnam
Systems architecture, model-output integration, data/API contracts, controlled replay, Mode Guard implementation and technical validation.