Maria Carolina Niño Rivera
About
Detail
Ingeniera de Sistemas
Bogota D.C., Colombia
Résumé
Jobs
verified_user
0% verified
- ASoftware Architect public Remote experienceAbogada Alexandra LozanoFeb 2024 - Jun 2026 (2 years 5 months)Architectural Evolution: Successful transition from a purely backend-focused role to a hybrid Software Architecture and AI Solutions Engineering position. Robust Backend Ecosystem: Design and structuring of relational database models from scratch and optimization of complex queries in SQL Server and PostgreSQL using NestJS (TypeScript) and Python. AI Orchestration and n8n: Implementation and design of multi-agent architectures and advanced agentic workflows using n8n to automate legal operational tasks and drastically reduce manual intervention. Intelligent Document Processing: Development and integration of OCR-based text extraction and validation pipelines with extensive language models (LLM) using Azure OpenAI Service and Google AI Studi
IconoiICONOI S.A.May 2016 - Feb 2024 (7 years 10 months)Optimización Extrema de Datos (Logro Crítico): Refactorización y diseño de procedimientos almacenados dinámicos en bases de datos relacionales, logrando reducir el tiempo de ejecución de un proceso crítico de negocio de 25 minutos a tan solo 6 segundos (99.6% de ahorro en procesamiento). Resiliencia Nacional (Caso Instituto Nacional de Salud): Diseño de la arquitectura de base de datos y backend en C# (.NET MVC) y Vue.js para la recolección de incidentes de salud pública. 1,514 Extreme Data Optimization (Critical Achievement): Refactoring and design of dynamic stored procedures in relational databases, reducing the execution time of a critical business process from 25 minutes to just 6 seconds (99.6% processing savings). National Resilience
Education
verified_user
0% verified
- Master's Degree in Artificial IntelligenceUniversidad Sergio ArboledaFeb 2022 - Apr 2024 (2 years 3 months)
- UEspecialización en ingeniería de softwareUniversidad Distrital FJCFeb 2018 - Mar 2019 (1 year 2 months)
- Ingeniera de sistemas public Remote experienceFundacion Universitaria Los LibertadoresAug 2011 - Jun 2016 (4 years 11 months)- Desarrollo y pruebas de aplicaciones Desarrollo utilizando C#, asp.Net, MVC .Net, capas. SQL Server, vue, javascript. Creación, mantenimiento, actualización y documentación de software.
Projects (professional or personal)
verified_user
0% verified
- AAuditorAbogada Alexandra LozanoJan 2026 - Jun 2026 (6 months)Auditor is a highly scalable enterprise orchestration engine that automates critical document analysis using agentic AI. Its success lies in a decoupled and dynamic architecture that reads evaluation rules directly from a PostgreSQL database, building adaptive prompts in real time without altering a single line of code. Furthermore, the system is completely model-agnostic, allowing for on-the-fly switching between different LLMs and dynamic adjustment of key parameters such as temperature based on the use case. Finally, thanks to its asynchronous, high-concurrency, multithreaded processing, it handles large volumes of data in parallel, guaranteeing structured outputs in clean, validated JSON schemas ready for direct integration with platfor
- LibrarySearch.AIPersonalFeb 2025 - Feb 2026 (1 year 1 month)I designed and implemented LibrarySearch.AI, a personal project focused on solving the problem of processing unstructured data—or text 'blobs'—in search systems. I built it in .NET 8 using a multi-project Clean Architecture, ensuring complete decoupling from the business logic. The workflow combines Generative Artificial Intelligence and transactional searches. In the Infrastructure layer, I implemented GeminiAiExtractionService to consume Google Gemini's LLM, configuring a strict structured generation scheme to force deterministic native JSON responses without conversational text. Once the AI cleanly extracts the entities (Title, Author, Keywords), a use case (MatchBooksUseCase) orchestrates the workflow by applying the Pattern Strateg
- Detection and classification of diabetic and hypertensive retinopathyUniversidad Sergio ArboledaFeb 2022 - Apr 2024 (2 years 3 months)For my Master’s thesis, I designed and implemented an advanced deep learning pipeline for the 'Detection and classification of diabetic and hypertensive retinopathy'. The architecture utilized a U-Net model specifically optimized to perform precise semantic segmentation of retinal fundus images, isolating critical vascular structures and biomarkers. Following the segmentation phase, I integrated an EfficientNet backbone as a feature extractor, paired with a custom classification layer to accurately categorize the severity and progression of both diabetic and hypertensive pathologies. This project demonstrates my ability to engineer complex, multi-stage computer vision pipelines, handle highly specialized medical datasets, and design high-fi
Awards
verified_user
0% verified
- CCum LaudeFeb 2022 - Apr 2024 (2 years 3 months)By academic average GPA 4.88/5.0 and honorary thesis 5.0
Publications
verified_user
0% verified
- IMorphological segmentation and level classification for diabetic and hypertensive retinopathy through ophthalmic imagesIngeniería y CompetitividadAug 2023 - Feb 2024 (7 months)The main objective of this research is to perform the segmentation and classification of retinal fundus images with diabetic and hypertensive retinopathy. A combination of a UNet convolutional network and a ConvNet was proposed for vessel mask segmentation and retinopathy classification, respectively. The classification process is based on ten defined classes, where values from 0 to 4 represent diabetic retinopathy and values from 5 to 9 correspond to hypertensive retinopathy. The approximate results for segmentation were Jaccard indices of 74%, F1 indices of 85%, and an accuracy of 96%, and for classification, an accuracy of 80%.