Francesc Coll
Sr Data Science Consultant · CLEARPEAKS
BIn My Own Words
LOCATION
Passionate about numbers, I have coursed statistics BARCELONA,
studies in Barcelona with two exchange programs in SPAIN
Lille (FR) and Sheffield (UK). The final thesis consisted
of developping a Statistical Software with JAVA.
My Data Science professional career was oriented to bu-
siness sector where I have acquired a pragmatic vision
of problem solving. Got deeper into complex machine
learning models, specially in Operational Research.
Work Experience
Sr Data Science Consultant
CLEARPEAKS2021 – Present
Classification model: Calculated the probability of failure for 120k water pipelines at Ag-
bar with an operational budget of 21M€. Applied XGBoost and AdaBoost on KNIME AP.
Demand Forecasting: Created a new time series model for sales forecast at BIC Graphic. It has been applied to 1k items to improve the purchasing and manufacturing processes. The multivariate time series model fits the data with exogenous factors like 'promotions' and 'out of stock'. Models used: SARIMAX on Azure ML Studio in Python.
Sr Data Science Consultant
FinTech StartUp2015 – 2021
Asset Management: Web solution for financial portfolio optimization. The model esti- mates the expected return and the variance (risk) from the historical stock prices. Af- terward, the solution gives a diversified portfolio that maximizes the return/risk ratio with an optimization algorithm. Models used: Mean-Risk Optimization in Python and NodeJs.
Professor
AMOS BUSINESS SCHOOL2016 – 2018
Applied Mathemathics: Teached subjects on marketing, finance, computer science and operations for 200 students from Bachelor to Master.
Data Science Consultant
AIR FRANCE / KLM2013 – 2015
Customer Insight • Portfolio Segmentation • LTV Modeling • Program Forecasting for upgrades & downgrades. Used Markow Chains in SAS.
Controller Internship
L'OREAL Luxe2013 – 2013
P & L • Dashboard & KPI
International Credit Risk Analyst
ONEY2007 – 2011
Credit Scoring: Developed a Scoring Model for credit card holders that was applied to 200k costumers/year. The observed risk was reduced by 30%. Model used: Logistic Regression in SAS..
Marketing Analyst
DELBARD2006 – 2007
Portfolio Manager • Data Broking • Portfolio Segmentation
Education
Master in International Business
IESEG BUSINESS SCHOOLMaster in International Business
University Degree
UNIVERSITAT POLITECNICA DE CATALUNYAUniversity Degree
Master's Degree
UNIVERSITE LILLE 1Master's Degree