Duong Kim Vinh
Duong Kim Vinh applies predictive modelling to your portfolio data continuously, so recommendations are ready when you reconnect — whether that's a hotel lobby, a co-working space, or an airport lounge with patchy signal.
Core Capabilities
The platform is designed to produce defensible, risk-adjusted recommendations rather than speculative signals. Each output is traceable to the data that informed it.
The engine processes historical and live market data to project probable outcomes across your current positions, updating as new data arrives.
Recommendations factor in asymmetric risk management, weighting downside exposure more heavily than upside potential to protect capital through market swings.
Portfolio concentration, correlation, and drawdown risk are consolidated into one dashboard, so you can assess your position without reconstructing it manually.
Methodology
Rather than ask for trust on faith, we explain each stage of the pipeline. You can audit the logic behind any given output at the point it is generated.
Market feeds, account positions, and macroeconomic indicators are ingested continuously and normalised into a common structure for analysis.
Trained models evaluate patterns across the normalised dataset, generating probability-weighted scenarios rather than single-point predictions.
Scenarios are filtered against your stated risk tolerance to produce a ranked recommendation, delivered with its supporting rationale.
Set up takes a short onboarding session to connect your accounts and define your risk parameters. From there, the analysis runs continuously in the background.