Javier Tarrio

💼 Professional experience

Agilytic

• Banking

Elaboration and deployment of scoring models for marketing purposes, based on product ownership and marketing engagement data of customers.

• Insurance Data extraction and transformation using SQL
• Food retail Elaboration of scoring models for product recommendation, based on most similar products and crosscorrelation between different customers. Optimization and enhancement of automatic periodic data transformation and ML processes.
• Interim management Deployment of a cloud data warehouse for analytic and predictive use of data. The mission included the study and redesign of the databases already in place, the deployment of the cloud infrastructure, with a configuration fit to the needs, and the building of ETL pipelines from data ingestion to report production.
• Cultural Sector – Client segmentation Based on several data sources, including ticketing information, developed a segmentation model to address clients in a more personalized way, also taking COVID effect into account. Iterations with the end-users was key in developing the most relevant and actionable segments. Technologies: Python for data analysis, modelling and visualisation (plotly, matplotlib), git for version controlling.
• Public sector – Supervised document classification Combining the scraping of thousands of websites and the usage of robust data pipelines and supervised modelling techniques, designed and built an effective Document Classification Engine leveraging an optimized AWS Cloud infrastructure in terms of cost and efficiency. Technologies: Python for data preparation & modelling, git for version controlling, AWS environment for industrialisation (incl. but not limited to S3 storage, Container Docker Repository, Batch, Aurora MySQL), terraform to deploy cloud resources.
• Automotive – Risk scoring Leveraging all available information at the subscription of a leasing contract, developed and deployed a comprehensive model to predict the risk and the severity of several types of incidents. He also shared in a very didactic way the results with the business to ensure adoption. Technologies: Python (with matplotlib and plotly in jupyter notebooks) for data analysis and visualization, TIMi for data preparation, modelling and deployment.
• Real estate – End-to-end reporting Developed a robust tool consolidating leads data from various websites into one data pipeline feeding the client’s CRM. Technologies: Python for data retrieval and preparation, AWS environment for industrialisation (PythonAnywhere), Power BI for data visualisation.

University of Helsinki, Finland

Université Libre de Bruxelles

• Lectured at theoretical physics master program.
• Supervised master students.

Post doctoral work

• Université Libre de Bruxelles, BE
• Universitat de Barcelona, ES
• Universiteit Utrecht, NL

Universidade de Santiago de Compostela, Spain

🦾 Certifications, trainings

Methodologies

• Data processing, analysis & visualization
• Data Pipelines industrialisation
• Numerical methods & analysis
• Machine learning methods
• Project Management
• Collaboration & management of international teams
• Infrastructure as Code

Software & programming

• Python (incl. Pandas, numpy, scipy, scikit)
• Spark DataFrame API and Databricks
• R
• SQL
• Tensorflow knowledge
• Anatella (ETL)
• Terraform

Certifications

• AWS Cloud Practitioner Certified
• Azure (AZ-900 certified)
• Databricks Certified Associate Developer for Apache Spark 3.0 (with python)
• Power BI Certified
• Azure Data Scientist Associate (ongoing)
• Terraform Associate

🎓 Academic credentials

• PhD in Physics Universidade de Santiago de Compostela, 2010

🇺🇳 Languages