Page Not Found
Page not found. Your pixels are in another canvas.
Sitemap
A list of all the posts and pages found on the site. For you robots out there is an XML version available for digesting as well.
Pages
A brief description about me
About me
Page not in menu
This is a page not in th emain menu
Posts
portfolio
Cognitive Architecture for ML Monitoring (CAMA)
LLM-based agent framework for autonomous ML model monitoring in production environments
Scanflow: MLOps Framework
Production-grade MLOps framework for ML workflow management, debugging, and reliability 
TADIL: Task-Agnostic Domain Learning
Novel continual learning algorithm enabling 6% performance improvement across domains without task labels
3D RAG for Healthcare Documents
Cost-efficient document parsing and retrieval at Cotiviti: semantic, lexical and graph retrieval (1.5k docs for $0.5)
Press: Cotiviti
KC-Agent: Dual-Process Cognitive Architecture
A dual-process cognitive architecture for efficient ML model improvement, published at IEEE COMPSAC 2026
Paper: arXiv
Adaptive Prompt Evolution (APE)
Autonomous prompt evolution for continual learning in diabetic retinopathy detection: 99% storage reduction, 96.9% accuracy
Paper: Springer Nature
TADILER: Continual Learning for Medical Imaging
Experience replay and zero-shot clustering for continual learning in diabetic retinopathy detection, VISAPP 2025 Best Student Paper Award finalist
Paper: VISAPP 2025
Improver Agent
First ‘Improver’ agent for automated ML model updates at SUNAT customs: 65% less manual intervention, real-time fraud detection
publications
Scanflow-K8s: Agent-based Framework for Autonomic Management and Supervision of ML Workflows in Kubernetes Clusters
Published in IEEE/ACM International Symposium on Cluster, Cloud and Internet Computing (CCGrid), 2022
Agent-based framework for autonomic management and supervision of ML workflows in Kubernetes clusters, published at IEEE CCGrid 2022.
Recommended citation: Bravo-Rocca, G., et al. (2022). "Scanflow-K8s: Agent-based Framework for Autonomic Management and Supervision of ML Workflows in Kubernetes Clusters." IEEE CCGrid 2022. https://doi.org/10.1109/CCGrid54584.2022.00047
Human-in-the-loop online multi-agent approach to increase trustworthiness in ML models through trust scores and data augmentation
Published in IEEE Computer Software and Applications Conference (COMPSAC), 2022
Human-in-the-loop online multi-agent approach to increase trustworthiness in ML models through trust scores and data augmentation, published at IEEE COMPSAC 2022.
Recommended citation: Bravo-Rocca, G., et al. (2022). "Human-in-the-loop online multi-agent approach to increase trustworthiness in ML models through trust scores and data augmentation." IEEE COMPSAC 2022. https://doi.org/10.1109/COMPSAC54236.2022.00014
Scanflow: Multi-graph framework for Machine Learning workflow management, supervision, and debugging
Published in Expert Systems With Applications, 2024
Production-grade MLOps framework for ML workflow management, debugging, and reliability published in top-tier journal.
Recommended citation: Bravo-Rocca, G., et al. (2024). "Scanflow: Multi-graph framework for Machine Learning workflow management, supervision, and debugging." Expert Systems With Applications. Impact Factor: 8.665. https://dl.acm.org/doi/10.1016/j.eswa.2022.117232
TADIL: Task-Agnostic Domain-Incremental Learning through Task-ID Inference using Transformer Nearest-Centroid Embeddings
Published in International Conference on Pattern Recognition (ICPR), 2024
Task-agnostic domain-incremental learning approach using transformer nearest-centroid embeddings for effective domain adaptation without task labels.
Recommended citation: Bravo-Rocca, G., et al. (2024). "TADIL: Task-Agnostic Domain-Incremental Learning through Task-ID Inference using Transformer Nearest-Centroid Embeddings." ICPR 2024. Kolkata, India. https://link.springer.com/chapter/10.1007/978-3-031-78110-0_22
PeruMedQA: Benchmarking Large Language Models (LLMs) on Peruvian Medical Exams - Dataset Construction and Evaluation
Published in arXiv, 2025
Benchmark of large language models on Peruvian medical exams, including dataset construction and evaluation.
Recommended citation: Bravo-Rocca, G., et al. (2025). "PeruMedQA: Benchmarking Large Language Models (LLMs) on Peruvian Medical Exams - Dataset Construction and Evaluation." arXiv. https://arxiv.org/abs/2509.11517
Experience Replay and Zero-shot Clustering for Continual Learning in Diabetic Retinopathy Detection
Published in International Conference on Computer Vision Theory and Applications (VISAPP), 2025
Novel approach combining experience replay and zero-shot clustering for continual learning in healthcare, shortlisted for Best Student Paper Award.
Recommended citation: Bravo-Rocca, G., et al. (2025). "Experience Replay and Zero-shot Clustering for Continual Learning in Diabetic Retinopathy Detection." VISAPP 2025. Porto, Portugal. https://doi.org/10.5220/0013128600003912
Feature Engineering for Agents: An Adaptive Cognitive Architecture for Interpretable ML Monitoring
Published in International Conference on Autonomous Agents and Multiagent Systems (AAMAS), 2025
This paper presents an adaptive cognitive architecture for interpretable machine learning monitoring in agent systems.
Recommended citation: Bravo-Rocca, G., et al. (2025). "Feature Engineering for Agents: An Adaptive Cognitive Architecture for Interpretable ML Monitoring." International Conference on Autonomous Agents and Multiagent Systems (AAMAS). Detroit, USA. https://dl.acm.org/doi/10.5555/3709347.3743552
Adaptive Prompt Evolution for Continual Learning in Diabetic Retinopathy Detection
Published in SN Computer Science (Springer Nature), 2026
Autonomous prompt evolution achieving 99% storage reduction while maintaining 96.9% accuracy in privacy-sensitive healthcare applications.
Recommended citation: Bravo-Rocca, G., et al. (2026). "Adaptive Prompt Evolution for Continual Learning in Diabetic Retinopathy Detection." SN Computer Science. https://link.springer.com/article/10.1007/s42979-026-05036-y
KC-Agent: A Dual-Process Cognitive Architecture for Efficient ML Model Improvement
Published in IEEE Computer Software and Applications Conference (COMPSAC), 2026
A dual-process cognitive architecture for efficient ML model improvement, published at IEEE COMPSAC 2026.
Recommended citation: Bravo-Rocca, G., et al. (2026). "KC-Agent: A Dual-Process Cognitive Architecture for Efficient ML Model Improvement." IEEE COMPSAC 2026. https://arxiv.org/abs/2608.02351
talks
Lenovo Press
Published:
Authored article for Lenovo Press on how autonomous vehicles use the TADIL algorithm to learn from changes in the environment.
ContinualAI Community
Published:
Invited talk for the ContinualAI community on TADIL, task-agnostic domain-incremental learning for the open world.
Xarxa RDI-IA Conference 2024
Published:
Poster on TADIL at the Xarxa RDI-IA Annual Conference, organized by AGAUR.
CVPR Workshop 2024
Published:
Extended abstract on continual learning in diabetic retinopathy at the LatinX in Computer Vision Research Workshop (CVPR 2024).
ICPR 2024
Published:
Oral presentation of TADIL at ICPR 2024, selected from among all accepted papers.
AAMAS 2025
Published:
Presentation of the CAMA cognitive architecture for interpretable ML monitoring at AAMAS, the premier conference on AI agents and multiagent systems.
Cotiviti Press
Published:
Press contribution highlighting the cognitive architectures being built at Cotiviti to help ML models adapt to real-world change.
teaching
MLOps and AI Systems Training
Industry Training, Barcelona Supercomputing Center & Lenovo, 2022
Led training sessions and knowledge transfer on MLOps, AI system reliability, and production ML deployment for industry partners and research teams.
AI Research Mentoring at Emory University
Research Mentoring, Emory Global Diabetes Research Center, 2023
Research mentoring and collaboration at Emory Global Diabetes Research Center focusing on continual learning applications for health and epidemiological research.