VaibhavMangroliya
Quantitative Developer | Risk Modeling | AI Systems
“Exploring how mathematical models break under extreme conditions — and how to build systems that account for it.”
Time-Series Forecasting (Risk Bounds)
Estimating uncertainty bounds via real historical distributions
Professional Trajectory
Bridging the gap between mathematical theory and engineering execution.
The Journey
I am a Mathematics Master's student who loves solving tough problems, whether it's pricing options, building ML models, or extracting insights from messy data. With ~2 years in fintech, I bridge the gap between mathematical theory and practical applications.
My journey into quantitative work began during my time at the National Stock Exchange of India. While building tools like the NAV calculator, Fair Value classification system, and XBRL financial data parsers, I kept running into the math underneath — pricing logic, valuation hierarchies, risk frameworks. That hands-on exposure is what sparked the pivot toward a more quantitative career path.
Now I sit at the intersection of quantitative finance and AI, pursuing my M.Sc. at the University of Luxembourg, alongside my work at the Luxembourg Institute of Science and Technology. What drives me? Honestly, I just love solving hard problems. Give me something messy and complicated, like figuring out how to price a tricky option or stress-test a portfolio, and I'm happy.
Education
M.Sc. in Mathematics
University of Luxembourg
Mathematical Modelling & Computational Sciences
09/2024 – Present
B.E. in Electronics & Telecommunication
Vidyalankar Institute of Technology
Grade: 1.4/1
08/2018 – 05/2022
Work Experience
Intern, ENVISION Unit (LEO Observatory)
Luxembourg Institute of Science and Technology (LIST)
- ▹Python-based environmental data validation, QA/QC pipelines, automated reporting.
Student Assistant, Dept. of Mathematics
University of Luxembourg
- ▹Preparation of technical documents and research materials using LaTeX.
Associate Systems Analyst
National Stock Exchange of India (NSE)
- ▹Regulatory compliance systems (Java/Spring Boot) for 2,700+ listed companies.
- ▹NAV calculation tool (Python, Oracle DB) automating Fair Value hierarchy classification.
- ▹XBRL parsing system transforming unstructured financial data into 23-table SQL schema. 40% error reduction.
Core Risk Systems
Interactive tools for stress-testing portfolios under extreme uncertainty. Built to explore real risk modeling concepts.
Portfolio Risk Engine
Real-time VaR & CVaR with fat-tail adjustments, multi-regime stress testing, and drawdown analysis.
Where Models Break
Standard models fail under extreme conditions. These visualizations demonstrate why risk engineering requires going beyond textbook assumptions.
Fat-Tail Distributions
Markets exhibit leptokurtic distributions. Standard Gaussian models dangerously underestimate extreme events.
Volatility Clustering
Large changes follow large changes. Mandelbrot's insight breaks the i.i.d. assumption.
Correlation Breakdown
Diversification vanishes during panics. All correlations → 1 when you need protection most.
Crisis Simulations
2008 & COVID: What actually happened vs what models predicted. The gap is catastrophic.
"The fundamental flaw of modern finance is substituting elegant but wrong mathematical assumptions for messy reality. The map is not the territory."
Gaussian Theory vs Empirical Reality
Decision Engine
A demo showing how raw risk signals — volatility spikes, correlation breakdowns, tail triggers — can be translated into concrete portfolio actions at different stress levels.
Select Scenario
1. Raw Signals
2. Risk Inference
3. Execution Engine
Quant Lab
Interactive mathematical tools for risk analysis. Every parameter is adjustable. Every chart updates in real time.
GBM Simulation
Geometric Brownian Motion paths under risk-neutral measure
Selected Systems
AI and ML models deployed as risk management tools. Each includes live interaction and multi-level explanation layers.
Pricing & Greeks
Black-Scholes-Merton model visualizing Option Payoffs and Greeks.
Call Option Value vs Intrinsic
Systemic Risk & Anomaly Detection
AI-driven anomaly detection for institutional risk monitoring and systemic protocol stress identification.
Open-SourceOpen-SourceProjects
Deployed repositories spanning quantitative modeling, machine learning, and environmental econometrics.
WHAT MODELS MISS IN REAL MARKETS
1. The Volatility Spike
A shock hits the system. Normal distribution models predict this event should happen once every 10,000 years, but it happened today. Implied volatility shoots up. Market makers widen spreads to protect themselves.
2. Correlation Breaks 1
As margins get called, funds are forced to liquidate everything. Equities, bonds, gold, crypto—the diversified portfolio fails. All correlations trend rapidly toward 1.0. There is nowhere to hide.
3. Tail Losses Materialize
The extreme left tail of the distribution curve realizes itself. This is why we engineer systems for the unexpected rather than the average. Risk is not about standard deviations; it's about what happens in the tail.
How I Think About Risk
Concise, high-signal principles that drive every system I build.
“Risk is tail exposure, not variance”
“Correlation breaks under stress”
“Models are approximations, not truths”
“Extreme events dominate outcomes”
System Architecture
How raw financial data is transformed into actionable risk intelligence through a 6-stage pipeline.
Market Data
Live feeds, tick data, macro indicators
Feature Eng
Volatility, momentum, risk factors
Risk Models
VaR/CVaR, Monte Carlo, GARCH
Decision Layer
Signal processing, regime detection
API Backend
FastAPI model serving & caching
UI Engine
Interactive risk visualization
Professional Endorsements
What managers and academics have said about my work.
"Vaibhav consistently stood out as a sharp and dependable professional. He showed a high level of ownership in his work, often handling critical modules with minimal guidance. Beyond his technical skills, Vaibhav is a collaborative team player with a professional attitude. I confidently recommend him for roles that require strong analytical thinking and problem-solving ability."
View on LinkedIn"Vaibhav approached mathematical challenges with a blend of intellect and methodical precision, delivering solutions of remarkable quality. He possesses remarkable analytical acumen—his ability to think critically and his strong ethics set him apart as a promising student."
View Full LOR"I can attest to his exemplary work ethic, adaptability, and quick learning abilities. Vaibhav's exceptional interpersonal skills and value as a team member, combined with his ability to seamlessly integrate new knowledge, speaks volumes about his versatility."
View Full LORInsights & Research
Why Normal Distribution Fails in Finance
Markets have fat tails. Assuming Gaussian returns is the most dangerous simplification in quantitative finance.
Building a High-Performance Risk Engine
How I designed a real-time VaR/CVaR engine from scratch using Python and FastAPI.
LSTM vs Classical ARMA Models
A deep comparison of deep learning vs statistical approaches for financial time-series forecasting.