Pregrado Ingeniería de Sistemas
URI permanente para esta colecciónhttp://hdl.handle.net/11634/192
Examinar
Envíos recientes
Tipo de ítem: Ítem , Conciliación de Pagos Inmediatos basada en Datos: Automatización, Trazabilidad y Analítica(Universidad Santo Tomás, 2026-10-08) Páez Ferreira, Carlos David; Santamaría Granados, Luz; Mendoza Moreno, Juan Francisco; Universidad Santo Tomás; https://scienti.minciencias.gov.co/cvlac/visualizador/generarCurriculoCv.do?cod_rh=0000348554; https://scholar.google.com/citations?user=1Eq9xg0AAAAJ&hl=en; https://orcid.org/0000-0003-3697-7852Instant payment systems settle individual instructions in seconds and are available twenty-four hours a day. Their architecture combines a central operator, which routes and validates instructions, with a set of participating entities that send and receive transactions. The irrevocable and continuous nature of these transfers shifts control to a later point after execution: the daily reconciliation, which verifies that what each participant reported, what the central processor recorded, and what was actually settled all describe the same set of facts. For the same business day, three sets of inputs with different origins, formats, and responsible parties come together, and decisions with direct monetary effects depend on how they intersect: calculating multilateral net positions, generating settlement messages, and sending responses back to each participant. Processing happens on machines without Internet access, which rules out remote analytics services and any dependencies that need to be downloaded at runtime.Tipo de ítem: Ítem , Guía Práctica para la Implementación de ISO/IEC 27001(Universidad Santo Tomás, 2026-10-06) Bermudez Nuñez, Samuel Ricardo; Contreras Ortiz , Martha Susana; Universidad Santo Tomás; https://scienti.minciencias.gov.co/cvlac/visualizador/generarCurriculoCv.do?cod_rh=0000901571; https://scholar.google.com.co/citations?user=L45gJqUAAAAJ&hl=es&oi=ao; https://orcid.org/0000-0002-7715-6420In the current technological landscape—defined by rising cyber threats and the accelerated digital transformation of organizations—safeguarding information assets has become a strategic pillar. However, significant gaps exist regarding the implementation of up-to-date international frameworks within specific business environments. Against this backdrop, this analysis of the ISO/IEC 27001:2022 standard addresses a genuine sectoral need arising from deficiencies in structured information security management. The absence of established processes considerably increases exposure to incidents that compromise operational continuity, institutional trust, and the protection of sensitive data (Alassaf & Alkhalifah, 2021). Grounded in an academic and disciplinary perspective, this research justifies its undertaking by synthesizing and systematizing the practical application of the ISO/IEC 27001:2022 standard within an organizational context. The study adds value to the field through the analytical identification of security gaps, the evaluation of updated controls (including the new controls in Annex A), and the proposal of conceptual guidelines aimed at risk mitigation. Thus, this monograph not only seeks to structure a threat prevention model but also establishes the theoretical and methodological foundations to drive continuous improvement processes and support informed decision-making (Andersson et al., 2022). The knowledge generated by this study directly benefits information security professionals and systems auditors in the productive sector by offering them an analytical framework to assess and strengthen their organizations. Furthermore, it serves as a reference for the academic community interested in management.Tipo de ítem: Ítem , Aplicación Móvil y de Escritorio POS Offline para la Empresa Full Colombia S.A.S.(Universidad Santo Tomás, 2026-10-06) De la Vega Vargas, Santiago; Soler Galindo, Harrizon Alexander; Universidad Santo Tomás; https://scienti.minciencias.gov.co/cvlac/visualizador/generarCurriculoCv.do?cod_rh=0002351384; https://orcid.org/0009-0002-6422-8429This work presents the results of an internship at Full Colombia S.A.S., a company based in Tunja that develops Full Suit, a point-of-sale and electronic invoicing solution. The work focused on the system's local application, built with Flutter, which operates offline on a SQLite database and synchronizes its data with the backend services once the network is restored. The desktop application was refactored, including a redesigned interface, fixes to pagination errors and form validation, and a new cashier shift module with an end-of-shift receipt and versioning of the local database schema. An Android version was developed, with two point-of-sale modules designed for small screens. A dynamic pricing system was implemented that lets the cashier set the final price and derives the taxable base, the discount and the tax through an inverse formula, coordinated with the web platform so that both applications produce the same value and the electronic invoice is accepted by the DIAN. Finally, the work was integrated into the team's SCRUM workflow, and synchronization was redesigned as independent steps that tolerate partial failures, together with exportable diagnostic logs. All four objectives were met, and the features are in operation at the businesses that use the system.Tipo de ítem: Ítem , Machine Learning en la Transformación Digital de las Telecomunicaciones: Aplicaciones y Desafíos(Universidad Santo Tomás, 2026-10-01) Gutiérrez Alvarado, Emerson Yezith; Castro Barinas, Manuel Leonardo de Jesus; Universidad Santo Tomás; https://scienti.minciencias.gov.co/cvlac/visualizador/generarCurriculoCv.do?cod_rh=0002116817Digital transformation has revolutionized the telecommunications industry, creating new challenges in process management and organization, as well as impacting network security. In this context, machine learning (ML) has emerged as a key technological solution. This article aims to investigate the influence of machine learning on the digital transformation of telecommunications, analyzing its primary applications and the challenges associated with its implementation. To this end, a narrative review of recent scientific literature was designed and supplemented with classic sources from databases such as Scopus, IEEE, Xplore, and ScienceDirect. The results reveal that Machine Learning significantly aids in improving predictive and optimization processes, network connectivity, and strengthening cybersecurity through real-time anomaly detection. However, the effective implementation of machine learning faces three major limitations: inadequate technological infrastructure, resistance to organizational change, and a shortage of specialized talent. The results demonstrate that machine learning is a key technology for digital transformation, but it depends on a combination of innovation, technological investment, and professional training.Tipo de ítem: Ítem , Sistema Web para Evaluación y Análisis del Índice GWI en Cuencas Hídricas(Universidad Santo Tomás, 2026-10-01) Casas Bello, Nicolás; Contreras Ortiz, Martha Susana; Caro Camargo, Carlos Andrés; Universidad Santo Tomás; https://scienti.minciencias.gov.co/cvlac/visualizador/generarCurriculoCv.do?cod_rh=0000901571; https://scholar.google.com.co/citations?user=L45gJqUAAAAJ&hl=es&oi=ao; https://orcid.org/0000-0002-7715-6420This undergraduate thesis addresses the lack of comprehensive technological tools for monitoring water resource management in rural watersheds under high anthropic pressure, given that current monitoring systems focus almost exclusively on physical and instrumental variables, leaving aside aspects of management, planning, and governance. To address this need, a modular web application was developed for recording, calculating, and analyzing the GWI (Green Watersheds Index) in rural watersheds, applying the XP agile methodology. System requirements were defined based on user stories gathered from key stakeholders in the water and environmental sector, and on guidance from the project's advisor and co-advisor, the latter being the author of the index's conceptual framework. The platform allows users, entities, projects, and watersheds to be managed, enables periodic evaluation of the 18 official indicators, and graphically displays the index's behavior by watershed and over time, including a public consultation module that promotes transparency in water governance. Tests carried out across the different user profiles validated the system's correct operation, confirming that the stated objectives were met and establishing a foundation for future enhancements aimed at monitoring water governance.Tipo de ítem: Ítem , Módulo de Gestión y Análisis Automatizado de Resultados Saber 11 y Saber Pro(Universidad Santo Tomás, 2026-10-01) Villamil Gonzalez, Juan Camilo; Suarez Rojas, Angie Lorena; Universidad Santo Tomás; https://scienti.minciencias.gov.co/cvlac/visualizador/generarCurriculoCv.do?cod_rh=0001829344The intervention in the Quality Assurance Directorate (DAC) is justified by the need to improve the current process for managing and analyzing the results of the Saber 11 and Saber Pro tests. However, the limitations of the existing system make timely access to information difficult and affect decision-making efficiency. Not addressing this situation could maintain technical dependence, cause delays in data analysis, and limit the use of key information for academic quality assurance processes. From a technical point of view, developing an automated document management module is proposed as an appropriate solution, since it allows structuring and optimizing existing processes, reducing usage complexity, and making it easier for DAC users to operate. This approach does not mean completely replacing the current system, but rather improving it through code reorganization, task automation, and integration with tools like Power BI. In this way, the proposed solution directly addresses the identified needs and aligns with the institutional objectives of continuous improvement. The expected impact focuses on optimizing the processes for managing and analyzing academic information, directly benefiting the DAC and the users responsible for institutional analysis. Likewise, a reduction in processing times is expected, along with less reliance on technical staff and better use of data for strategic decision-making. In terms of scope, the internship is focused on the development and implementation of the proposed module within the established timeframe, considering the available technological conditions and access to information provided by the department. Finally, from a training perspective, the internship allows applying knowledge in software development, databases, and information analysis in a real context, strengthening both technical skills and abilities related to problem-solving and teamwork. This contribution is understood as a complementary result, with the main goal being to help improve a relevant institutional process.Tipo de ítem: Ítem , Información Clínica en Español: Cómo Aprovecharla sin Poner en Riesgo la Privacidad(Universidad Santo Tomás, 2026-09-30) Ospina Leguizamón, Daniel Mauricio; Piracoca Arcos, Jhon Alexis; Universidad Santo Tomás; https://scienti.minciencias.gov.co/cvlac/visualizador/generarCurriculoCv.do?cod_rh=0002450348The growth of unstructured clinical information, such as medical records, clinical notes, and narrative reports, has created new challenges for its organization, analysis, and use in the healthcare sector. In this context, natural language processing (NLP) and language models have become key tools for extracting, structuring, and protecting the relevant information contained in medical texts. This popular science article was developed through a descriptive, qualitative literature review of fifty academic, technical, and regulatory sources, with the aim of explaining the importance of unstructured data in healthcare, describing the main applications of NLP in this field, and analyzing its most significant challenges, such as model accuracy, patient privacy, and the de-identification of sensitive information. The findings show that much of the most valuable clinical information resides in free text, that transformer-based techniques and language models can extract it with growing accuracy also in Spanish and that de-identification and interoperability are indispensable conditions for its ethical use. It is concluded that unstructured data, when handled securely, represents a real opportunity to strengthen research, support clinical decision-making, and modernize health information systems.Tipo de ítem: Ítem , Ciencia de datos e inteligencia artificial en la predicción de eventos cardiovasculares en pacientes hipertensos: una revisión integrativa de la literatura(Universidad Santo Tomás, 2026-07-10) Núñez Niño, Daniel Enrique; Suárez Rojas, Angie Lorena; Universidad Santo Tomás; https://scienti.minciencias.gov.co/cvlac/visualizador/generarCurriculoCv.do?cod_rh=0001829344Diseases of the cardiovascular system are the leading cause of mortality in the world. Arterial hypertension is the main cardiovascular risk factor that can be avoided. In reducing deaths from these diseases, the use of innovative technologies in this region is of great importance. This popular article emphasizes the data of the Tunja region, a municipality that has been considered the health area in data, and disease data of Tunja, and therefore, the forecasts of the use of Data Science and Artificial Intelligence in the analysis of the prediction of cardiovascular events in hypertensive patients are presented. The review of studies and literature that have been published in databases, and that currently generate great impact in the area of health technologies, has been systematic, and the studies have been critically analyzed, in the use of machine learning tools and neural networks. The application showed that models such as Random Forest, XGBoost, and cylindrical neural networks of long-term and short-term memory, presented more than 88% efficiency in the prediction of acute myocardial infarction, event v and/or cardiovascular death in hypertension companies. The incorporation of data science tools in open health systems provide awareness about prevention resources and reduce the essentials of medicine.Tipo de ítem: Ítem , Inteligencia Artificial Generativa: ética, sesgos y sostenibilidad(Universidad Santo Tomás, 2026-07-09) López González, Juan Felipe; Vergel Camacho, John Jairo; Vela Beltrán, Diego Alejandro; Universidad Santo Tomás; https://scienti.minciencias.gov.co/cvlac/visualizador/generarCurriculoCv.do?cod_rh=0001977029; https://scholar.google.com/citations?user=o-A8ASAAAAAJ&hl=es; https://orcid.org/0009-0003-1809-4033Background: Generative Artificial Intelligence (GAI) represents one of the most profound and accelerated technological transformations in recent decades. Systems such as GPT-4, Claude, Gemini, DALL-E, Midjourney, and Stable Diffusion have demonstrated unprecedented capabilities to redefine the boundaries between human and automated production. However, this disruption unfolds upon historical structures of inequality, power concentration, and informational asymmetries that GAI models absorb, reproduce, and amplify. Methodology: This research constitutes a systematic literature review of high-impact scientific literature, executed under the PRISMA 2020 protocol. The study analyzes 130 indexed academic sources published between 2016 and 2026, selected from an initial universe of 104,660 records across four international databases: IEEE Xplore, EBSCO Research, Scopus (Elsevier), and Web of Science (Clarivate). The investigation examines the influence of cognitive biases, training data quality, and environmental impact, with a sectoral emphasis on health and justice and a specific projection toward the Latin American context. Results: The study identified six primary categories of bias: representation, measurement, evaluation, implementation, LLM-specific, and historical-structural. In healthcare, evidence shows that models trained on electronic records often confuse differential access to care with actual clinical needs. In justice, seemingly neutral variables serve as proxies for race, reproducing discrimination under technical objectivity. Regarding mitigation, while techniques like data rebalancing, adversarial learning, and frameworks (NoBIAS, EU AI Act) exist, their effectiveness is moderate and involves significant trade-offs; improving intergroup equity often results in a decrease of up to 10 percentage points in global accuracy. Sustainability: "Green AI" strategies—such as model pruning, quantization, and knowledge distillation—can reduce energy consumption by 13% to 115%. However, only 23% of Green AI studies involve industrial partners, revealing a gap between technical knowledge and practical adoption. Conclusions: Ethical and sustainable GAI development requires simultaneous interventions: binding regulatory frameworks, independent audits, and algorithmic transparency. For Latin America and Colombia, the challenge lies in overcoming "algorithmic transport failure"—the dependence on models from the Global North that fail to capture local linguistic, cultural, and socioeconomic nuances. Reversing this vulnerability requires not only technical investment but the construction of a unique epistemic perspective that positions the region as an active subject in the contemporary technological order.Tipo de ítem: Ítem , Dashboard Interactivo para la Analítica y Visualización de Resultados Saber Pro en la USTA(Universidad Santo Tomás, 2026-07-06) Montes Patiño, Christian Alberto; Soler Gómez, David Gerónimo; Universidad Santo Tomás; https://scienti.minciencias.gov.co/cvlac/visualizador/generarCurriculoCv.do?cod_rh=0001343613; https://scholar.google.es/citations?user=31nUOG8AAAAJ&hl=esThe relevance of this project is based on the urgent need of the Quality Assurance Department at Santo Tomás University to optimize strategic decision-making by moving from a manual, decentralized, and error-prone handling of microdata from the Saber Pro tests to a centralized and automated institutional analytics system. Implementing this solution is justified as it reduces current operational rework and, at the same time, is expected to have a significant medium- and long-term impact on university management; on one hand, it will give deans and curriculum committees a handy tool to identify gaps in generic competencies and design timely pedagogical improvement plans before official audits, and on the other, it will set up a standardized and scalable data infrastructure that will serve as the methodological basis for future predictive analytics studies and measurement of added value through relational cross-referencing with historical Saber 11 databases. The execution of this project is essential to modernize the mechanisms of academic auditing at Universidad Santo Tomás, moving from a static data scheme toward a culture of institutional analytics. Meeting the main objectives will make it possible to deliver concrete solutions and tools that democratize access to information, allowing deans and researchers to explore multidimensional indicators in real time to design effective teaching strategies and support high-quality accreditation processes.Tipo de ítem: Ítem , Gestión del Conocimiento en Proyectos Tecnológicos Ágiles en Sistemas GyG S.A.(Universidad Santo Tomás, 2026-07-10) Márquez Camargo, Jorge Alberto; Guío Ávila, Henry Alfonso; Universidad Santo Tomás; https://scienti.minciencias.gov.co/cvlac/visualizador/generarCurriculoCv.do?cod_rh=0001519211; https://scholar.google.com/citations?user=gqWnDVQAAAAJ&hl=es; https://orcid.org/0000-0003-1343-4302This internship at Sistemas GyG S.A. aims to solve a real problem the company faces: when they want to try out new technologies or features for their financial system SIIF, they don’t have a clear and quick way to validate if those ideas will actually work before investing time and money to develop them fully. This means that sometimes they go ahead with projects that later don’t meet expectations, causing wasted resources and delays. My job will be to create prototypes and quick tests that let us know from the start whether a technology is worth it or not, as well as set up clear testing processes to make sure whatever is developed works well from the beginning. I’ll also document everything we learn in each project so that knowledge isn’t lost when people change roles or leave the company, but is available to everyone. The impact of this work will be tangible both for the company and for the sector. GyG Systems will be able to have a clearer analysis of which technologies to evaluate or adopt in the development of new features, which will contribute to more informed decision-making aligned with the needs of the main product. This is important because competition in the financial sector is growing, especially with the arrival of tech startups that quickly adapt to market changes. By documenting processes and results in an organized manner, the company will also be able to build a knowledge base that team members can easily reference, allowing them to learn from previous experiences and continuously improve development practices. For Boyacá and Colombia, this kind of initiative helps strengthen national companies competing in the tech sector, promoting talent development and innovation in the region.Tipo de ítem: Ítem , Simulación de Gemelos Digitales para el Control de Calidad del Aire en Espacios Públicos(Universidad Santo Tomás, 2026-07-22) Salamanca Rodríguez, Fabian Camilo; Contreras Ortiz, Martha Susana; Universidad Santo Tomás; https://scienti.minciencias.gov.co/cvlac/visualizador/generarCurriculoCv.do?cod_rh=0000901571; https://scholar.google.com.co/citations?user=L45gJqUAAAAJ&hl=es&oi=ao; https://orcid.org/0000-0002-7715-6420The focus of this article is on showing how air quality simulations based on Digital Twins, integrating low-cost simulation technologies like Wokwi and Microsoft Azure's cloud infrastructure, can predict and visualize pollution levels in critical public spaces, such as hospital settings and urban parks. The proposed approach is based on the design of the Internet of Things (IoT), where the Wokwi simulator mimics the behavior of ESP32 microcontrollers paired with particle and gas sensors, while Microsoft Azure serves as the central ecosystem for processing, storing, and analyzing the data collected through Azure IoT Hub and Azure Digital Twins. The most significant conclusion reached shows that precise two-way synchronization is achieved between the virtual simulation environment and the digital model in the cloud Corporates make it easier to design urban mitigation strategies without having to spend on expensive physical hardware during early validation stages. The proposed application is a scalable, replicable, and highly efficient way to plan smart cities with environmental sustainability and public health in mind, using high-fidelity synthetic data to maximize tech decision-making.Tipo de ítem: Ítem , Módulo web para la gestión de sistemas de información, programación de backups y reportes de mantenimiento del Hospital Universitario San Rafael de Tunja(Universidad Santo Tomás, 2026-07-15) Sanchez Marquez, Santiago; Piracoca Arcos, John Alexis; Universidad Santo Tomás; https://scienti.minciencias.gov.co/cvlac/visualizador/generarCurriculoCv.do?cod_rh=0002450348This internship thesis describes the development and implementation of a web module for the San Rafael University Hospital of Tunja (HUSRT), integrated into the existing HUSRT-Biomédica system. The main objective was to centralize and automate the management of information systems, backup scheduling, and maintenance reports to address the operational gaps identified in the Information Technology Department (ITD). Problem: The hospital lacked a unified system inventory; backups were managed manually without a visible calendar or alerts; access control was insufficient; and the service desk was not integrated with the equipment inventory. Developed Solution: A module was built featuring: Information systems parameterization (CRUD with Angular, Express, and MariaDB). Automated backup scheduling with 7 configurable frequencies (daily, weekly, biweekly, monthly, quarterly, semi-annually, annual) and automatic yearly occurrence generation. Interactive visual calendar with color coding by status (Scheduled, Completed, Not Done, Failed), tooltips, management modals, and Excel export. Reactive alerts via a bell with badge and overlay panel, with auto-regeneration of the next backup upon completion. Granular access control based on role and assigned systems, including the new SYSTEMUSER role. Service Desk integration with the equipment inventory, allowing cases to be linked to equipment and generating structured maintenance reports upon case closure. Methodology: An agile approach was used with 27 epics and 69 user stories, validated by the company tutor. Technologies included Angular 17 (standalone), Node.js + Express, Sequelize (ORM), and MariaDB, with PrimeNG for the UI and JWT for authentication. Results: The module was approved and deployed to production, reducing backup monitoring time from several hours per week to less than 30 minutes daily, eliminating forgotten backups, improving information security (aligned with Colombian Law 1581 of 2012), and providing complete traceability of equipment maintenance.Tipo de ítem: Ítem , Sistema de seguimiento a graduados multifacultad para la Universidad Santo Tomás Seccional Tunja(Universidad Santo Tomás, 2026-07-10) Molano Cubillos, Richard Eduardo; Vela Beltrán, Diego Alejandro; Universidad Santo Tomás; https://scienti.minciencias.gov.co/cvlac/visualizador/generarCurriculoCv.do?cod_rh=0001977029; https://scholar.google.com/citations?user=o-A8ASAAAAAJ&hl=es; https://orcid.org/0009-0003-1809-4033The code was refactored to scale the graduate management system of the School of Systems Engineering—originally designed for a single faculty—to a multi-faculty environment at Santo Tomás University (Tunja Branch). This involved structuring the data model and adopting a modular architecture to improve software traceability and maintainability, while retaining the original Angular and NestJS frameworks and implementing Dockerization to facilitate deployment. Additionally, the interfaces were refactored, and a corresponding database architecture was proposed, alongside the implementation of the interfaces and logic required to achieve the desired scalability.Tipo de ítem: Ítem , Ética de la inteligencia artificial, marcos globales y sus limitaciones(Universidad Santo Tomás, 2026-07-10) Cuervo Martinez, Duban Fernando; Soler Galindo, Harrizon Alexander; Universidad Santo Tomás; https://scienti.minciencias.gov.co/cvlac/visualizador/generarCurriculoCv.do?cod_rh=0002351384; https://orcid.org/0009-0002-6422-8429International governance is undergoing an extraordinary shift as large language models (LLMs) transition from simple theoretical concepts to crucial geopolitical and regulatory players. At present, a small group of dominant corporations leads the global rollout, highlighting the shortcomings of current regulatory systems in addressing underlying ethical concerns. Utilizing a systematic literature review under the PRISMA protocol, this study provides a critical examination of how dominant firms OpenAI, Anthropic, Meta, Google, and Microsoft manage transparency, internal governance, and corporate accountability within their commercial systems. Data collection spanned the specialized Scopus and IEEE Xplore databases across eight structured search strategies; after screening 959 initial records, a definitive corpus of 69 articles published between 2017 and 2026 was curated. The primary findings expose a profound disconnect between corporate ethical rhetoric and actual operational practices. Moreover, ambitious legal initiatives such as the European Union AI Act exhibit structural blind spots when attempting to oversee models that are inherently opaque and probabilistic. A critical bias also emerges in current literature: the systematic marginalization of the Global South, a deeply concerning trend given the swift, unmonitored integration of these tools across Latin America. This paper concludes with actionable recommendations to anchor corporate ethics firmly within LLM development, tailoring the analysis to the distinct regulatory landscapes of Colombia and the broader region.Tipo de ítem: Ítem , USTA Gallery: Plataforma web para gestión y exhibición del semillero de artes(Universidad Santo Tomás, 2026-07-08) Cardozo Diaz, Oscar Javier; Castro Barinas, Manuel Leonardo de Jesús; Universidad Santo Tomás; 0009-0003-1767-6517USTA Gallery is a web platform developed as a undergraduate thesis project by Oscar Javier Cardozo Díaz at Universidad Santo Tomás, Tunja campus, under the supervision of Ing. Manuel Leonardo Castro Barinas. The problem: the university's arts semillero lacked a digital channel. Artwork was only visible at in-person exhibitions, with no participation records or data to justify the group's continuity before university authorities. The solution: a full-stack platform with a NestJS/Bun backend, PostgreSQL database managed with Prisma ORM (21 tables), and a React 19 frontend with TypeScript and Tailwind CSS. The entire system runs on Docker Compose over a Linux virtual machine provided by the university's TIC group, accessible at gallery.santototunja.edu.co. The system features three roles — student, teacher and administrator — JWT authentication stored in an HttpOnly cookie, and email verification via OTP. It includes 10 functional modules: artwork management with a PENDING → APPROVED / REJECTED workflow, a public gallery, cultural event management with a group invitation system, and a real-time class attendance module added as a scope extension beyond the original objectives. Testing was conducted in two layers: 95 unit tests with Jest at 100% pass rate, and 125 integration requests with Newman covering 9 modules. Result: all 4 specific objectives met and the platform fully operational in production. Main limitation: the semillero was inactive during development. Future work includes multi-tenancy support, WebSockets notifications, and an institutional analytics module.Tipo de ítem: Ítem , Inteligencia Artificial Generativa: Ética, Sesgos y Sostenibilidad(Universidad Santo Tomás, 2026-07-09) Vergel Camacho, John Jairo; López González, Juan Felipe; Vela Beltrán, Diego Alejandro; Universidad Santo Tomás; https://scienti.minciencias.gov.co/cvlac/visualizador/generarCurriculoCv.do?cod_rh=0001977029; https://scholar.google.com/citations?user=o-A8ASAAAAAJ&hl=es; https://orcid.org/0009-0003-1809-4033Background: Generative Artificial Intelligence (GAI) represents one of the most profound and accelerated technological transformations in recent decades. Systems such as GPT-4, Claude, Gemini, DALL-E, Midjourney, and Stable Diffusion have demonstrated unprecedented capabilities to redefine the boundaries between human and automated production. However, this disruption unfolds upon historical structures of inequality, power concentration, and informational asymmetries that GAI models absorb, reproduce, and amplify. Methodology: This research constitutes a systematic literature review of high-impact scientific literature, executed under the PRISMA 2020 protocol. The study analyzes 130 indexed academic sources published between 2016 and 2026, selected from an initial universe of 104,660 records across four international databases: IEEE Xplore, EBSCO Research, Scopus (Elsevier), and Web of Science (Clarivate). The investigation examines the influence of cognitive biases, training data quality, and environmental impact, with a sectoral emphasis on health and justice and a specific projection toward the Latin American context. Results: The study identified six primary categories of bias: representation, measurement, evaluation, implementation, LLM-specific, and historical-structural. In healthcare, evidence shows that models trained on electronic records often confuse differential access to care with actual clinical needs. In justice, seemingly neutral variables serve as proxies for race, reproducing discrimination under technical objectivity. Regarding mitigation, while techniques like data rebalancing, adversarial learning, and frameworks (NoBIAS, EU AI Act) exist, their effectiveness is moderate and involves significant trade-offs; improving intergroup equity often results in a decrease of up to 10 percentage points in global accuracy. Sustainability: "Green AI" strategies—such as model pruning, quantization, and knowledge distillation—can reduce energy consumption by 13% to 115%. However, only 23% of Green AI studies involve industrial partners, revealing a gap between technical knowledge and practical adoption. Conclusions: Ethical and sustainable GAI development requires simultaneous interventions: binding regulatory frameworks, independent audits, and algorithmic transparency. For Latin America and Colombia, the challenge lies in overcoming "algorithmic transport failure"—the dependence on models from the Global North that fail to capture local linguistic, cultural, and socioeconomic nuances. Reversing this vulnerability requires not only technical investment but the construction of a unique epistemic perspective that positions the region as an active subject in the contemporary technological order.Tipo de ítem: Ítem , Módulo de gestión y control de equipos del área de sistemas para el Hospital Universitario San Rafael de Tunja(Universidad Santo Tomás, 2026-07-09) Ramirez Mancilla, Juan Miguel; Piracoca Arcos, Jhon Alexis; Universidad Santo Tomás; https://scienti.minciencias.gov.co/cvlac/visualizador/generarCurriculoCv.do?cod_rh=0002450348Hospital Universitario San Rafael de Tunja relied on an information system for recording and tracking its technology assets that, as the hospital's infrastructure grew, revealed significant functional limitations: it lacked role-based access control, provided no way to schedule preventive maintenance, offered no traceability of changes made to each device, and failed to produce consolidated reports for decision-making. This internship aimed to improve the hospital's HRCATCH system by developing a technology asset management module, applying the Scrum agile methodology, in order to centralize inventory administration and support the IT department's operational decisions. The development was carried out in four phases: analysis of existing processes and requirements through meetings with IT staff; design of a three-layer architecture consisting of an Angular 19 frontend, a Node.js/Express REST API, and a MariaDB database managed with Sequelize; development of the functional modules for inventory, equipment lifecycle, preventive maintenance, and role-based access control differentiated by user profile (AdminSistemas, TécnicoSistemas, and UsuarioSistemas); and validation of the system through functional testing against the specified requirements. The result was a functional module that centralizes technology inventory management, automates preventive maintenance scheduling, generates Excel and PDF reports, and incorporates a technological obsolescence index calculated from four weighted variables (equipment age, manufacturer support validity, frequency of corrective maintenance, and parts availability), replacing the previous subjective criteria with a process grounded in verifiable data. The project demonstrates that rigorous application of software engineering principles, combined with proper institutional requirements gathering, can produce technology tools of concrete, measurable value for the healthcare sector.Tipo de ítem: Ítem , Desafíos éticos y normativos de la Autenticación biométrica con Inteligencia Artificial(Universidad Santo Tomás, 2026-07-10) Villanueva Arias, Juan Felipe; Contreras Ortiz, Martha Susana; Universidad Santo Tomás; https://scienti.minciencias.gov.co/cvlac/visualizador/generarCurriculoCv.do?cod_rh=0000901571; https://scholar.google.com/citations?user=fxmDbqoAAAAJ&hl=enThe main issue addressed is the growing vulnerabilities of biometric authentication systems in the face of emerging capabilities in artificial intelligence, which in turn creates unprecedented ethical and regulatory challenges that compromise the integrity, privacy, and trust in these security systems.Tipo de ítem: Ítem , Integración en InvuPos con Ecosistemas Externos: Contabilidad, Delivery y Gestión de Personal(Universidad Santo Tomás, 2026-07-07) Blanco Sandoval, Jesus Esteban; Piracoca Arcos, Jhon Alexis; Contreras Ortiz, Martha Susana; Universidad Santo Tomás; https://scienti.minciencias.gov.co/cvlac/visualizador/generarCurriculoCv.do?cod_rh=0002450348; https://scholar.google.com/citations?hl=en&user=04JqgRoAAAAJ; https://orcid.org/0009-0001-9382-1451This document presents an internship project focused on building a backend integration ecosystem between INVU POS and external delivery and accounting platforms, specifically Rappi, PedidosYa, Siigo, and QuickBooks. The project covered requirements gathering, architecture design, implementation of functional modules, and integration testing, with the goal of automating processes, improving system interoperability, and strengthening operational efficiency. As a result, the project delivered solutions capable of synchronizing menu, order, and accounting information, providing technical and operational value to the platform.