Duc Hien VU

– SOLVENCII started as a side-project (Varathon) in early 2020, aiming to build Web Interface for existing ALM software. – In 2023, we started building our own ALM models and algorithms (mostly in Python) and leveraging Cloud Computing & Web Assembly for deployment. – Specially designed for actuarial and finance purposes, our SaaS software platform combines powerful features for actuarial data analysis, ALM modeling, Solvency 2 & IFRS17 reporting, and much more. – We provide modern, high-performance, flexible web applications that remain intuitive and accessible. – Our actuarial models are transparent and accompanied by detailed documentation, auditable calculations, and can be accessed through our integrated Python console and Excel-like web-based spreadsheet. – We also offer tailored consulting and training services (with or without our SaaS).

– Master 280 - Statistical and Financial Engineering (ISF): Course: Introduction to Asset-Liability Management – Master 218 - Insurance and Risk Management: Course: Risks, Values and Performance

– Strategic Asset Allocation & ALM Prospective Analysis (Solvency II - pillar II) – ORSA, PRIIPS, Stress Tests, Structured Product Studies, etc. – Studies of Risk Neutral & Real World Scenarios Generators – Strategic Asset Allocation

– Regulatory Monitoring Team (Solvency II - Pillars I & III) – BE, SCR, S2 ratio production and analysis on a quarterly basis. – Analysis of Change (VA C2C, AoC for S2 Own Funds) – Technical studies (ESG, Profit Sharing Strategies, Investment Strategies, Competitive Rate Modeling, cyclical lapses ...), – Process Automatisation & Optimisation (for input preparation, model validation, mirror modeling, result analysis, QRTs ...) – Economic Own Funds' analysis based on contracts' VIF, linking Solvency II own funds with MCEV, IFRS17 visions ...

– My main mission was to develop an actuarial model for Major Losses Reserving at AXA CS. – In this model, the Major Losses Threshold is determined according to some techniques derived from the Extreme Values Theory. – Once this threshold calibrated, the model will separately estimate the IBNER amounts (using an individual method that I've named Individual Chain-ladder) and the IBNyR amounts (thanks to a frequency-severity approach and an aggregated method). – I developped a corresponding desktop application (with R and Shiny package). – Here is a post of mine about deploying an user-friendly desktop application with R, Shiny, and a web browser: http://littleactuary.github.io/blog/Web-application-framework-with-Shiny/ – I also participated to the 3rd quarter reserving (Q3) & forecasting (F3) process and reinsurance analysis.

– I was in charge of improving and speeding-up a CAT/Atypical modeling & reinsurance pricing tool (with R and C++, using Rcpp package). – Here's a post of mine about speeding-up R programs with Rcpp package in R: http://littleactuary.github.io/blog/connecting-R-and-Cpp-with-Rcpp/ – I also helped restructure the program, added new functionalities and prepared the integration of the R core into a web platform.

– The main mission was to calibrate different dental claim distributions (with SAS) and to deploy a tool for measuring the pricing impact of a dental cap (with Excel/VBA). – I have also conducted a research on behaviors of health insurance policyholders before their retirement.