Nathan Azoulay

Case Files

Projects, studies and systems organised around the question they make visible — with their level of evidence stated plainly.

CASE 01
Applied machine learning

Machine Learning for Insurance Reserving

Automatic detection of claims requiring closure

A supervised-learning study developed at SADA Assurances to support the prioritisation of open non-life insurance claim files likely to require closure.

Master’s dissertation · industry internship
Evidence structure
Question

Prioritise review without automating the claim decision.

The project focused on a practical reserving question: when many claim files remain open, can a model help identify those that deserve earlier review? The proposed role was triage and prioritisation, not autonomous closure.

Method

Define the target, prepare historical data, compare models.

The workflow linked target definition, historical claims preparation, feature engineering and supervised-model comparison. Evaluation considered precision, recall, specificity, AUC and balanced accuracy so that different error patterns remained visible.

Evidence

A decision-support case, assessed through more than one score.

The dissertation received 17/20. The case file is structured around operational interpretation: how a predictive signal could support a review queue, what its scope is, and why a good model score is not sufficient on its own.

What it demonstrates: connecting a business decision, a defined target and an interpretable evaluation framework.

CASE 02
Interactive system

Dilemma Royale

Iterated prisoner’s dilemma simulator

An interactive game-theory application that turns the iterated prisoner’s dilemma into a visual, replayable system of agent behaviour, strategy tuning and simulation.

Independent application
CASE 03
Quantitative finance

Volatility Surface Calibration

Black–Scholes, implied volatility and local volatility

An exploratory study of option-pricing calibration against market data, examining implied volatility, local-volatility exercises and the gap between a model surface and observed prices.

Exploratory technical notebook
CASE 04
Quantitative market study

Brent Oil Study

A quantitative study of Brent oil data, retained as a case file for its market-data reasoning, time-series interpretation and technical reporting structure.

Technical study
CASE 06
Professional data practice

Iterato

Early applied data work linked to residential construction-cost prediction, combining programming and data-management support in a practical business context.

Applied experience
CASE 07
Research proposal

AI Content Disclosure, Brand Authenticity & Engagement

Young consumers in Spain · fashion, beauty and lifestyle marketing

A quantitative, scenario-based survey experiment examining whether disclosure of AI-generated social-media content changes engagement intention through perceived brand authenticity.

Draft research proposal · no results presented

Do you have a question that needs a clearer model, method or interface?

Let’s discuss