Scaling agentic AI systems in production for financial services

Important: register on the external event website is required for admission.

In this Silicon Valley edition, Nu’s AI Core team, the centralized AI research and engineering organization, will pull back the curtain on how machine learning systems are operationalized at scale. The speakers will bypass surface-level concepts to dive deep into real production code, infrastructure trade-offs, and architectural constraints. Designed to be intimate and conversational, the evening prioritizes deep technical dialogue, peer connection, and high-density networking.

The session will cover:

• The infrastructure, data engineering, backend orchestration, and intelligence layer behind production-grade systems;
• Spanning predictive models trained on proprietary financial data;
• Autonomous agentic tool-calling pipelines;
• Custom simulation environments used for offline policy evaluation.

Who should attend:
This event is curated for ML Engineers, Senior Software Engineers, Tech Leads, and Systems Architects with a minimum of 5+ years of hands-on experience in core ML systems, large-scale backend, or distributed infrastructure. The session is specifically tailored for practitioners focused on model optimization, training efficiency, high-throughput cloud architecture, and the data pipelines required to deploy and orchestrate predictive and generative frameworks at scale. This space is designed for seasoned builders looking for deep architectural discussions on AI systems and peer-to-peer networking rather than surface-level product pitches.

Agenda:

• 5:00 PM – Welcome Reception & Networking (Food and soft drinks served)
• 5:30 PM – Opening Remarks/Intro
• 5:40 PM – Tech Talk 1, by Aman Gupta (AI Core, Nu)
• 6:10 PM – Tech Talk 2: Effective quantization of Muon optimizer states, by Shao Tang (AI Core, Nu)
• 6:40 PM – Panel Discussion + Audience Q&A
• 7:30 PM – Happy Hour & Open Mixer (Beer, wine, and refreshments)
• 8:30 PM – Event Conclusion