Six research lines, seven active projects, and contributions spanning healthcare, industry, security, and society.
In2Lab is dedicated to the study and application of intelligent methods for complex information problems. Our work addresses both fundamental research questions and real-world applications, with a strong emphasis on collaboration with clinical, industrial, and academic partners across Colombia and internationally. We publish in top-tier venues and actively share code and datasets.
Research on the design and development of intelligent information systems using artificial intelligence, including machine learning, deep learning, pattern recognition, signal and image processing, computer vision, recommender systems, intelligent agents, and natural language processing to extract knowledge, automate tasks, and support complex decision-making processes.
Research on methods and technologies for collecting, managing, processing, analyzing, and visualizing large volumes of data, including data mining and supervised and unsupervised learning, to discover patterns, generate knowledge, and provide evidence-based support for organizational decision-making and strategic planning activities.
Research on the design and development of software architectures and distributed systems, including microservices, cloud-native platforms, high-performance computing, resilient architectures, and cybersecurity mechanisms, to enable scalable, efficient, reliable, and secure operation of complex computational services and applications across diverse environments.
Research on models, methods, and standards for ensuring data quality and interoperability across heterogeneous systems, while addressing data management, information exchange, ethical principles, privacy, security, and responsible governance in the collection, integration, sharing, and use of large-scale data across organizations.
Research on strategies, models, and technologies for digital transformation and intelligent organizations, including technology adoption, digital maturity assessment, process automation, and robotic process automation, to improve organizational capabilities, operational efficiency, service delivery, and innovation in public and private organizations.
Define a conceptual and methodological framework based on the people, processes, and technology approach to support strategic planning for data governance in public institutions of higher education.
This project proposes the design of an AI-based co-adaptive control system for myoelectric transradial prostheses, aimed at improving users’ interaction with and continued use of the prosthesis. The approach combines user adaptation and system adaptation, allowing the control strategy to progressively adjust to changes in the user’s myoelectric signals, movement patterns, and interaction context.
This project focuses on developing predictive models to identify hospitalized patients at increased risk of readmission. Using clinical and demographic data, machine learning techniques can be applied to characterize patterns associated with subsequent hospital readmission and estimate individual patient risk.
Development of multimodal machine learning models that integrate clinical and chest radiographic data to classify tuberculosis lineages. The approach aims to exploit complementary information from both data modalities to improve classification performance and provide a more comprehensive characterization of tuberculosis case
This project aims to characterize the serum phospholipid profile across distinct stages of AD progression — from presymptomatic to symptomatic phases — using both a triple-transgenic mouse model (3xTg-AD) and human familial AD patients. By mapping phospholipid signatures at defined time points, the study seeks to identify early lipid biomarkers associated with disease onset and progression. These signatures will be correlated with neuroinflammatory and histopathological markers in the hippocampus to establish mechanistic links between peripheral lipid dysregulation and central pathology. Building on these findings, candidate phospholipids will be encapsulated in liposomes to assess, in vitro, the effects of both pathological and protective lipid signatures on astrocyte and microglial activity in co-culture with neurons. Together, this work aims to advance the understanding of phospholipid-driven neuroinflammation in AD and to evaluate liposome-based platforms as potential therapeutic or diagnostic tools for cognitive decline and dementia.
Development of machine learning models and continuous learning strategies for the automatic detection of fraudulent transactions in virtual debit and credit card systems.
For a complete list of publications associated with these projects, see the Publications page or our GrupLAC profile.