Barcelona MLOps #3: Agents and Evals
📅 July 24th
📍 Barcelona

Note: We are at full capacity but join the waiting list, there are always last minute drops.

We’re excited to bring the Barcelona MLOps community together again for our third edition — this time focused on agentic workflows and evaluation strategies in production ML systems.

🎙️ Speakers

Agostino Calamia - Autonomous Systems with Reproducible Outcomes: Agent Validation for Reliable Automation

Agostino will explore how agent validation frameworks—such as multi-agent consensus models, cross-validation protocols, and redundancy techniques—can improve reliability, traceability, and reproducibility in autonomous systems. Learn how these methods support dependable automation and reduce operational risks.

Ashley Duque Kienzle - Evaluating ML in the Agentic Era

Ashley will dive into the challenges and opportunities of evaluation in modern ML systems powered by agents. Her talk will explore how traditional metrics fall short, and what new approaches are emerging to assess performance, safety, and usefulness in agentic workflows.

Vista Crew - Agentic flow for SEO Ad Title Generation
Jack Lin: Principal Data Scientist
Oleguer Vallbe Aris: Principal Analytics Engineer
Omar Elbadrawi: Engineering Manager, Machine Learning

Jack, Oleguer and Omar will share a practical case study on building an agentic workflow to generate Google Shopping ad titles. This talk highlights key lessons learned, challenges faced, and real-world insights from deploying agent-based solutions in a high-scale marketing automation context.

Thanks to Prosus, OLX Group and Vistaprint for supporting the community.

Barcelona MLOps 3: Agents and Evals