Mathieu Thélot

The programmatic Company is a startup that distributes its clients' advertisements through ad buying platforms.

Descriptive Statistics and Business Approach:

User behavior data exploration in relation to advertisements

Conversion rate study based on standard and retargeting strategies

Campaign time modeling

Anomaly detection in storage or during advertising campaigns

Mathematical Approach:

Simulation of a bidding system used by Facebook or other platforms

Pure Business and Operational:

Media Trader

Implementation of interactive dashboard normalization (Microsoft Power Bi)

Creation of a module for aggregating data for Excel reports used by internal sales teams

Predictive Model:

Deployment and benchmarking of prediction algorithms.

Creation of a model for predicting click-through rates.

Some technologies used:

Python, numpy, pandas, scikit-learn,

SFR: Software evolution - improve ASMM processing chains - automated deployment (Shell - Java -

Oracle - SQL - Jenkins)

Data popularization: creation of a web mockup dedicated to data processing and big data methods and references

Wind production forecast: forecast of wind production more than 3 days ahead

Climespace: Customer consumption forecast model based on historical data. Forecast and anticipation

of the future, taking into account customer arrivals/departures

Ecare: predictive maintenance of connected thermostats - detection of insulation anomalies - IoT - Data collection - Inter-model performance comparison - Time series analysis - pre-processing - unit tests - Industrialization (AWS)