● Developed a sentiment analysis model for movie reviews by manipulating non-traditional data,performing preprocessing, normalization, and TF-IDF vectorization.● Tokenized the reviews and trained a custom word embedding using Word2Vec.● Achieved an 81% accuracy and an 89% ROC curve in detecting positive and negative opinions.● Developed a Machine Learning model for vehicle classification by training and evaluating differentmodels, including Multi-layer Perceptron, LeNet, and ResNet50, on a pre-built dataset of 196 classes.Achieved an 83% accuracy in predicting the type of car in images, with ResNet50 performing the best.● Developed a Machine Learning API system that receives images, predicts their content using theResNet50 convolutional neural network model, and achieves a 93% prediction score. Implemented APIendpoints, utilized a Redis database, and built the system with Docker and Docker Compose.Conducted stress tests with Locust to assess API capabilities and limits.● Developed a credit risk scoring model by obtaining and analyzing data using specialized libraries(Matplotlib/Seaborn). Preprocessed the data through feature selection, outlier detection/handling,feature imputation/normalization/standardization, and categorical feature encoding. Created abaseline model using randomized search and cross-validation techniques with Scikit-learn in Python,achieving a training ROC AUC score of 74% and a validation score of 70%.● Built an ELT Data Pipeline for e-commerce data from 2016-2018: obtained information from CSV filesand APIs, stored it in a database, loaded it into a data warehouse, created SQL queries to extract newdata by linking tables, and analyzed the relationship between holidays and purchase order dates usingPython and visualization libraries (Matplotlib/Seaborn).

Design and development of the auto insurance purchase system.● Collaborated within a Scrum team, contributing to backend development for microservicesimplementation.● Directly participated in designing and implementing highly scalable solutions.● Successfully integrated payment gateways for various insurance providers

● Train Large Language Models (LLMs) using Reinforcement Learning from Human Feedback (RLHF).● Enhance the accuracy of major LLMs worldwide.● Contribute to the development of LLMs from top global companies like Google, OpenAI, Meta, and others.● Compare outputs of various LLMs under training.● Improve LLM responses for code generation, Q&A, and other objectives.● Handle implicit code execution: Develop necessary code for prompts requiring additional code routines for execution.● Run test prompts to identify and fix model failures during the Q&A process.● Create prompts for training datasets, focusing on model weaknesses identified by customers.● Contribute to LLM training for major development companies

Mexican startup dedicated to the marketing of manufacturing industry machines for entrepreneurs.● Innovated by developing new e-commerce tools to enhance functionality.● Ensured optimal performance by maintaining internal company platforms.● Enhanced tools, both new and existing, through thoughtful design and improvements.

Streaming platform and video-analysis full automatic.● Executed script implementation for automating data retrieval using Node.js.● Regularly obtained and updated data using a cron schedule.● Ensured data integrity and efficiency through data cleaning and maintenance.