VIGIL-V1
Financial risk intelligence, built progressively.
An ML-based financial transaction risk-scoring engine designed to evolve from transaction-level assessment towards richer behavioural and relationship-based intelligence.
VIGIL-V1 is being developed as an ML-based financial transaction risk-scoring engine. Its initial focus is transaction-level risk assessment, with behavioural intelligence forming an important part of its product direction and graph-based relationship analysis representing a longer-term vision.
Risk hides in volume and in connections.
Institutions handle more transactions than any team can review by hand. Checks that look at one record at a time can miss patterns that only emerge across many accounts, or over time.
From transaction to explained score.
Transactions
Transaction activity enters the model. Current development focus.
Behavioural context
Activity is assessed in the context of account history. Product direction.
Relationships
Connections between entities add wider context. Longer-term vision.
Risk assessment and explanations
A risk assessment with the factors behind it, intended for analyst review. Design intent.
Development status
Transaction intelligence
Assess individual financial transactions using machine learning.
Behavioural intelligence
Add account-history context to identify unusual changes and patterns.
Relationship intelligence
Explore connections between entities to uncover network-level risk patterns.
Explainable intelligence
Make risk assessments understandable and reviewable by analysts.
VIGIL-V1 is under development. Capabilities described here reflect design intent and may change before release.
