Expert Data Scientist

• Feb 2023 - Present

Inteligencia artificial

• Mar 2018 - Feb 2023

Team management, business development and strategy definition. Recommendation engine for fashion and groceries: transformers for customer behaviour, product embedded representation and similarity search. Computer vision: product classification, object detection, image segmentation, image captioning and image similarity. NLP: texts classification, BERT fine tune, NER models, synonym detection and chatbots. Team management, business development and strategy definition. Recommendation engine for fashion and groceries: transformers for customer behaviour, product embedded representation and similarity search. Computer vision: product classification, object detection, image segmentation, image captioning and image similarity. NLP: texts classification, BERT fine tune, NER models, synonym detection and chatbots.

Business Manager Data Scientist

• Sep 2012 - Mar 2018

Main tasks: ▪ Team management ▪ Business development and client relationship management ▪ Definition of internal methodology for different types of projects as data and text mining, chatbots development, based on CRISP-DM and Agile Methods ▪ Participation in recruiting processes Industry: ▪ Classification model of claims via email whit 60 different categories, using NLP techniques and Deep Learning algorithms, such as CNN ▪ Segmentation of customers of a fuel company in order to achieve personalized actions/services obtaining the knowledge of each individual through the analysis of their movements stored in a Hadoop framework Banking: ▪ Development of a semantic search engine for the customer area of the bank that allows queries about data using expressions in natural language. The engine was built using deep learning algorithms working on Tensorflow ▪ Analysis of customer feedback in online banking in order to describe the main topics of the comment and the sentiment and emotion. Different text mining techniques were applied, LDA and CTM. For sentiment and emotion an algorithm was developed in order to determine if a feedback is positive or negative and if has one of the eight main emotions ▪ Obtaining the main attributes that allow the customers' experience. Result analysis in order to detect the most relevant levers for improving satisfaction and recommendation. Insurance: ▪ Development of a chatbot to manage employees’ problems with company usernames and passwords. The chatbot was developed using deep learning techniques, word embedding and BiLSTM-CRF, to recognize the main concepts of the user utterances Telco: ▪ Construction of a predictive churn model for telco customers in order to improve loyalty plans. The project was developed using machine learning capabilities over structured and unstructured data. Utilities sector ▪ Development of an automated classifier of customer calls in order to help the call center agents to solve the customer claims. Show less

Consultant Data Scientist

• 2008 - Nov 2012

Utilities sector▪ Gas and electricity customer clustering using unsupervised machine learning techniques▪ Construction of a predictive churn model using consumption and claims in order to help marketing areas to improve customer loyalty▪ Development of cross selling and up selling models for improving commercial actions▪ Support in ETL (Extract, Transform and Load) of data▪ Support in reporting and information visualizationTelco sector ▪ Development of a translator between SAS and SQL Teradata using as input modeling flows created in SAS MinerInsurance sector▪ Customer clustering using unsupervised machine learning techniques▪ Lifetime value calculation for auto and house insurance customers Show less

Becario

• 2008 - 2009

- Proyectos internos:Adaptación de una herramienta para la auditoría de datos desarrollada en Matlab para hacerla compatible con bases de datos Oracle. Creación de una herramienta de generación de variables sintéticas desarrollada en C#.