From AI models to AI systems: agents, multiagents, orchestration (II)
22 October 2026 | 09.05h to 10.00h
We continue with a second session focused on showcasing the potential of Agentic AI in real-world environments. Through four use cases, we will explore its application across diverse domains, ranging from drinking water network management and audiovisual content editing to software development and the automation of specialized functions such as DevOps operations.
Participants:
-
Andreu Giménez Larred, AI and Data Engineer, AVENTEC
-
Miquel Garcia Guerrero, Post-Production Manager, Gestmusic
-
Álvaro López Antequera, R&D firmware developer, HP
-
José Julián Ariza Valderrama, CEO, Connecting Tech People
Presenter:
Speakers:
ANDREU GIMÉNEZ LARRED
AI and Data Engineer, AVENTEC
Andreu Giménez is an AI Engineer and Data Engineer at Aventec SL, where he develops AI models and data architectures applied to the urban water cycle. He holds a BSc in Computer Engineering and an MSc in Data Science from UPC (FIB), and served as a teaching assistant for the CBI4AI course at ESADE, in collaboration with CERN.
AI applied to drinking water quality management: prediction of chlorine decay in the Vic network.
Ensuring water quality requires a global view of the behavior of the distribution network. This view is only possible by combining hydraulic simulation of the network with Artificial Intelligence models capable of interpreting the results. Our project aims to develop models that determine the optimal dosage of chlorine in chlorination stations, so that all points in the network remain within the quality limits established by the regulations.
MIQUEL GARCIA HERRERO
Post-Production Manager, Gestmusic
Miquel Garcia is a Post-Production Manager in the audiovisual industry, working at Banijay Iberia through Gestmusic. He specializes in overseeing post-production processes and drives innovation in the production of audiovisual formats. His work focuses on integrating new technologies and optimizing workflows, with the aim of exploring new ways of producing that support and enhance creativity. In a dynamic media environment, he collaborates closely with creative and production teams to deliver high-quality content for diverse audiences.
When AI Works as a Team: Agents, Multi-Agent Systems, and Audiovisual Storytelling with KAI
This session presents KAI, an AI-powered audiovisual storytelling platform that combines multimodal analysis, Large Language Models (LLMs), and multi-agent architectures to transform large volumes of audiovisual content into structured, searchable, and editable stories. Through a real-world television production use case, we will demonstrate how different AI agents collaborate to analyze content, generate metadata, identify narrative elements, and assist editorial decision-making, while integrating directly into professional post-production workflows.
ÁLVARO LÓPEZ ANTEQUERA
R&D firmware developer, HP
PhD in Industrial Engineering, with a Master’s in Automation and Robotics and more than 11 years of experience at HP, evolving from Firmware Developer to Firmware Architect. Strong background in firmware, industrial systems, automation, and robotics, with previous research experience at CSIC. Currently focused on architecture leadership and actively promoting data and AI adoption to improve engineering efficiency, scalability, and innovation.
From NLP Models to an Operational AI System for Automated Bug Triage
Large Format Printing firmware development produces high volumes of complex issues across multiple products and specialized teams, making manual triage inefficient and error-prone. This work presents an AI-driven system that automates bug triage using natural language processing and machine learning. The solution processes Jira data, predicts the responsible team, assigns issues automatically, and analyzes attachments to detect missing information and duplicates. An SVM-based stacking approach achieved the best performance, reaching 77% accuracy. Integrated at scale, the system reduces manual effort and cuts resolution time by over 30%, demonstrating how applied AI can deliver measurable impact in industrial software engineering environments.
JOSÉ JULIÁN ARIZA VALDERRAMA
CEO, Opsphere
José Julian Ariza is a Cloud Architect and DevSecOps specialist with more than 20 years of international experience leading technology transformation initiatives across multiple industries. His expertise includes cloud architecture, platform engineering, applied artificial intelligence, observability, and large-scale automation. Alongside his professional work, he actively participates in technical training and knowledge sharing, helping organizations adopt AI in a practical, secure, and business-oriented way. He have worked on global high-availability platforms, large-scale e-commerce systems, and production environments where AI is already part of daily operations.
