I’m Olver, and I’m passionate about building applications that make life easier for people and businesses using Full Stack and Machine Learning.
Projects Download CVWalmart records weekly sales between 2010 and 2012 across 45 stores. Data processing pipelines and time series forecasting models (ARIMA and SARIMA) are developed to predict demand, following the CPMAI methodology. The project also includes a backend application that exposes the trained models to generate predictions and query performance metrics.
Backend App Backend App Repo ML App Repo ML App Doc Walmart DatasetOlist E-commerce records multiple purchase orders, so data quality was evaluated to identify potential Machine Learning applications. Extraction and preprocessing pipelines were developed to build a consolidated dataset and explore analysis opportunities, following the methodology CRISP-DM. Google Cloud Storage was used to extract the data and BigQuery to preprocess it by joining multiple tables.
ML App Repo ML App Doc Olist DatasetCorrectly classifying materials requires identifying the appropriate recycling bin, a process that can be slow and difficult to scale when considering all bin colors. A notebook was developed to train CNN models, including the use of pretrained models, to classify materials according to the color of the recycling container.
Notebook Repo Integrated DatasetA Chifa restaurant needed a mobile version of its web ordering application. A mobile app was developed while preserving most of the design details from the web version. Integration with Getnet was also considered.
Mobile App RepoThe Saucache Campus of the University of Tarapacá is quite large, so users sometimes wonder how to get to a specific location. A system was developed to help users navigate using a map, a search tool, and a route planner, showing directions to both outdoor areas and building interiors. It also accounts for different walking and driving routes available across the campus.
Mobile App Repo Frontend App Repo Backend App RepoSecurity is one of the main concerns in residential communities. Therefore, a Panic Button Mobile Application was developed to send alarm notifications to preferred neighbors. Additionally, a notification service was integrated into the Backend Application, allowing users to both receive and create alerts.
Mobile App Repo Frontend AppThe Master’s Program system is a platform for managing postgraduate students’ academic activities, which originally relied on external Supabase services. For security reasons, prior to deployment, it was necessary for the data to be managed locally on the University’s servers. The services used included user authentication, file management, and email sending. Once the adaptation was completed, the system was dockerized and deployed on local infrastructure.
Auth & Email App Repo File Management App RepoI’m motivated to collaborate on meaningful projects and build real-world solutions.