ABOUT

---

Senior Expert Data Scientist with over 10 years of combined experience in both industry and academia, I am a seasoned professional in the eld of retail. From my academic background to my real world contributions, I have always sought to bridge the gap between theoretical knowledge and practical solutions. My professional motivations are rooted in the search for transformative change. Over the years, I have successfully led and contributed to various projects that leveraged my background in automation with computer vision, visual search and others. My journey has seen me collaborate in the national project of pricing where I effectively leveraged IBM Cloud Pak for Data to deliver results that have had a tangible impact. This ability to seamlessly blend academic rigor with real-world application has allowed me to not only stay at the forefront of industry advancements, but also consistently deliver measurable value to the organizations I've been a part of.

TECHNICAL TOOLKIT:

PROFESSIONAL EXPERIENCE

• Pricing Project for Furniture Category in Collaboration with MIT.

  This pricing project for the furniture category emerged as a cutting-edge collaboration between CENIC and the Massachusetts Institute of Technology (MIT). The initiative focused on optimizing pricing strategies to enhance pro tability in the national furniture market. Web Application Deployment: The nal outcome of the project was implemented and deployed on an intuitive web application, allowing users to access and utilize the new pricing strategies effectively. National Scope: The project addressed the furniture market on a nationwide scale. Protability Objective: The primary goal of this project was to increase the pro tability of the furniture category within an ambitious range spanning from 5% to 20%. This substantial enhancement in pro tability was achieved through strategic implementation of new pricing strategies.

  Tools Utilized: NumPy, Pandas, Scikit-Learn, Random Forest, Jupyter Notebook Server, IBM Cloud Pack, IBM CPLEX, SQL, DBeaver, Google Docs.

• Pricing Project for Clothing Category with MIT-Inspired Strategies and Advanced Models.

  This pricing project for the clothing category builds upon the suggestions from the Massachusetts Institute of Technology (MIT) and incorporates novel approaches to optimize pricing strategies. The project involved the development of cutting-edge models, including Feedforward Neural Networks (FNN) and Long Short- Term Memory (LSTM) networks, along with the application of unsupervised algorithms to group demand time series (Time Series Clustering). The project's scope extends nationally, catering to the clothing market across the country.

  Tools Utilized: Keras, TensorFlow, Azure Data Studio, SQL Server, Matplotlib, Seaborn, regression (parametric and non-parametric), Bayesian and non-Bayesian methods, Time Series Analysis and Forecasting, short-term forecasting.

• Business Case: Integration of GANs, Computer Vision, and Social Media Data Analysis in Fashion Design and Trend Forecasting.

  This business case proposes an innovative strategy for the fashion industry by combining advanced technologies such as Generative Adversarial Networks (GANs), Computer Vision, and social media data analysis. The objective is to revolutionize how clothing is designed and fashion trends are forecasted, enabling our company to stay at the forefront of the industry and make more informed and strategic decisions.

  Tools Utilized: PyTorch, OpenCV.

• Developing FER Model for In-Store Analysis Using Transfer Learning in TensorFlow and PyTorch.

  Led the development of a Facial Expression Recognition (FER) model for in-store analysis, aimed at understanding customer reactions and emotions. Implemented transfer learning techniques with Xception and Inception architectures in both TensorFlow and PyTorch frameworks.

  Tools Utilized: Google Colab.

• Computer Vision Automation.

  Led the conceptualization and execution of a Computer Vision-driven automation system, aimed at streamlining internal company processes. Developed a classi er using transfer learning with the MobileNet network in TensorFlow. The classi er was subsequently integrated into a API and is used by a a bot.

  Tools Utilized: SCRUM.

• Visual Search Demo Development on GCP.

  Led a team of four individuals in the creation of a visual search demo for the company, leveraging code from former colleagues. This project involved implementing the visual search functionality on Google Cloud Platform (GCP) to enhance image search and recognition capabilities.

  Utilized technologies such as Cloud Source Repositories, Docker, Object detection, semantic segmentation, Git, Linux, SCRUM and Flask to ef ciently deploy the demo.

• Analysis and Enhancement of Statistical Methodology in Data Analytics and Supply Chain.

  Played a pivotal role in reviewing and enhancing the statistical methodology employed in the data analytics and supply chain. Focused on analyzing and rectifying the existing methodology used for measuring business trials, this project involved identifying shortcomings and implementing new statistical tools to enable more accurate and reliable assessment of business trials.

  Tools Utilized: Hypothesis testing, regression (parametric and non-parametric), Bayesian and non-Bayesian methods, advanced statistical algorithms.

• Development of Unsupervised Algorithm for Time Series Clustering in Judicial Statistics Area.

  Played an essential role in implementing an innovative unsupervised clustering algorithm in the realm of judicial statistics. This project aimed to address the complexity of ef ciently allocating human resources based on patterns in timeseries data. I devised a solution that effectively grouped similar time series, facilitating strategic human resource allocation. The initial version of the algorithm was created in SPSS, later migrated to R to harness its advanced data analysis and visualization capabilities.

  Tools Utilized:Microsoft Oficce.

• Text Classi cation Model for National Census of Federal Justice Impartation.

  Led the development of a sophisticated text classi cation model as part of the "National Census of Federal Justice Impartation". This project aimed to pre-classify federal offenses based on information reported by jurisdictional bodies. The text classi cation model utilized natural language processing techniques to analyze and categorize the textual descriptions of reported offenses. By applying machine learning algorithms and feature engineering, the model achieved accurate and ef cient pre-classi cation of diverse federal crimes, enhancing the data analysis process for the census.

  Tools Utilized:Microsoft SQL Server.

• Served as a dedicated professor at the Faculty of Science, National Autonomous University of Mexico (UNAM), imparting knowledge and fostering academic growth among students. I designed and conducted a diverse range of courses, including Probability, Bayesian Statistics, Mathematics for Biology and Physics, Seminars on Actuarial Applications, Algebra, Stochastic Processes, Mathematical Analysis, Differential Equations, and Risk Theory. Through these courses, I aimed to cultivate students' understanding and application of mathematical and statistical concepts in various elds.

EDUCATION

---

• PhD in Mathematics

  National Autonomous University of Mexico, CDMX

  Expected Graduation: AUG, 2024.

  2012-2017

• M.S. in Science.

  National Autonomous University of Mexico, CDMX

  Graduated: JUNE, 2012.

  2010-2012

• Bachelor of Mathematics

  National Autonomous University of Mexico, CDMX

  Graduated: JAN, 2010.

  2006-2009

• Bachelor of Actuarial Science

  National Autonomous University of Mexico, CDMX

  Expected Graduation: AUG, 2024.

  2005-2007

GENERAL SKILLS

---

• Collaborative Team Player, Problem Solving, Effective Communication, Adaptability, Creativity, Solution-Oriented, Afable, Teamwork.

LANGUAGES

---

• English (fluent) and Spanish (native)

SOFTWARE DEVELOPMENT

---

• Apolonio v1. A graphical interface to teach a generalization of the circle, ellipse, parabola and hyperbola. Laboratory of scienti c computation of the National Autonomous University of Mexico. (Matlab): JAN 2007- DEC 2010.

  Tools Utilized: Matlab.

LECTURES

---

• Reconstrucción booleana de materiales porosos, invited talk in the DÉCIMA SEMANA INTERNACIONAL DE LA ESTADÍSTICA Y LA PROBABILIDAD, Puebla, Puebla. JUNE 2017.
• Una aplicación de la Teoría Ergódica en la búsqueda de Google in the XLII CONGRESO NACIONAL DE LA SOCIEDAD MATEMÁTICA, Zacatecas, Zacatecas. Oct 2017.

INTERESTS

---

• Exploring emerging trends in AI, machine learning, and automotive technology.
• Contributing to open-source data science projects.