Problem Description

The Servicio de Administración Tributaria (SAT) managed vast data volumes across multiple databases with complex naming conventions. Searching for and understanding tables and fields required the manual consultation of scattered data dictionaries (Excel files), which led to significant delays and operational errors in data access.

Solution Description

I developed a Web Application for Data Exploration and Cataloging. The tool functions as a centralized, dynamic data dictionary, enabling employees to quickly and intuitively search for and retrieve contextual information for any entity (database, table, or field).

Result Description

The implementation eliminated the reliance on external files, drastically reducing exploration time and improving the efficiency and accuracy in managing structured data across all departments.

Technologies Used

Problem Description

The internal restructuring of the AGCTI (General Administration of Communications and Information Technologies) information scheme and the launch of the New Analytical Platform (NPA) required a clear and unified communication strategy. There was a risk that SAT employees would be unaware of the operational changes and new analytical tools, which would hinder platform adoption and the utilization of the new data infrastructure.

Solution Description

I developed an informative, high-impact Landing Page, deployed internally using an IIS server. The site was designed to be the central information hub for the new data scheme, highlighting the benefits, structure, and key procedures of the NPA in collaboration with the Analytical Intelligence Area.

Result Description

The implementation of the landing page guaranteed effective and timely internal communication regarding the strategic changes. This facilitated a smooth adoption of the New Analytical Platform by SAT personnel and served as a permanent consultation repository for the new operational data model.

Technologies Used

Problem Description

The newspaper's content team faced a slow, manual publishing process that relied on the creation and manual formatting of HTML code for every news story. This created bottlenecks in production, consumed valuable time from the Copy Writer, and limited the ability to respond quickly to breaking news.

Solution Description

I developed a comprehensive and highly intuitive News Management Web Platform (CMS). The application integrated Artificial Intelligence (Gemini API) to automatically generate the HTML code and formatting for the news stories based on the text entered by the Copy Writer. This included publishing both previews and full articles.

Result Description

The implementation achieved complete automation of the publishing process, which drastically reduced news delivery times. The Copy Writer could focus on content quality instead of technical formatting, significantly increasing NativeBo's operational efficiency.

Technologies Used

Problem Description

The SAT was undergoing a drastic technological migration process which led to operational difficulties and a high dependency on technical support for employees to carry out their daily activities. The Analytical Intelligence Area needed an efficient solution to scale training and reduce the need for continuous assistance.

Solution Description

I developed a comprehensive Web Platform for Training Course Management. This digital solution allowed the administrative area to manage the entire course lifecycle, from content publication, attendance tracking, and feedback collection, to personnel progress monitoring.

Result Description

The platform scaled the SAT's training capacity, reducing dependency on continuous support and accelerating the adoption of new technologies. This resulted in a significant improvement in the operational efficiency of the internal staff.

Technologies Used

Problem Description

The design team faced the challenge of selecting and validating large volumes of visual resources efficiently. Manually verifying the authenticity, content (text and celebrities), and quality of images was a slow, error-prone process, in addition to presenting a growing risk of using manipulated or AI-generated images (deepfakes) in publications.

Solution Description

I developed a Visual Content Analysis Web Application that automated the resource validation process. The solution integrated AWS Rekognition for face detection and analysis, celebrity recognition, and text detection, along with the NVIDIA Deepfake Detection service to identify fake or AI-generated images. The tool served as a selection assistant for the designers.

Result Description

The implementation reduced the legal and editorial risk associated with using unverified content. Furthermore, it optimized the designers' workflow by automating validation, ensuring a faster, safer, and higher-quality resource selection for all company pages.

Technologies Used

Problem Description

In a large university, the role of the prefect—responsible for verifying schedule compliance and faculty attendance—was straining and inefficient. It required multiple people to patrol facilities recurrently, resulting in a high operational cost and slow response to absences.

Solution Description

I developed a Monitoring Web Application that automated attendance verification. The solution was based on training a computer vision model using Azure Custom Vision. This model detected faculty presence via images captured by cameras, sending the analysis results immediately to the centralized web application.

Result Description

The application eliminated the need for recurrent physical rounds, reducing operational strain and allowing prefects to manage attendance centrally and in real-time. This increased the efficiency and accuracy in monitoring faculty schedule compliance.

Technologies Used

Problem Description

NativeBo faced a high demand from clients who required a considerable number of advertising Landing Pages. The manual creation process and coordination between designers, copywriters, and developers was slow and difficult to scale, generating bottlenecks that negatively impacted delivery times.

Solution Description

I developed a Landing Page Creation Web Application that streamlined the workflow. The tool allowed users to select pre-defined sections from a database. Crucially, the application used the Gemini API to automatically generate the corresponding content (text) and styles (CSS), dynamically inserting them into the selected structure.

Result Description

The application became an essential tool that reduced the production time of Landing Pages to record time. This highly efficient workflow for the multidisciplinary team allowed the company to scale its response capacity to high demand and serve a broader catalog of clients quickly and profitably.

Technologies Used