SF AI

AWS Agentic AI Summit - June Edition

Wed, Jun 24 · 4:00 PM PDT

Location not specified yet

Register on the AICamp Event Website is REQUIRED for admission.

From Code to Production - Building Production-Ready AI Agents on AWS

AWS Agentic AI Partner Showcase (Fifth Edition) features experts from AWS, Saviynt, Deepgram, Zendesk, 4Minds, who will demonstrate how to build production-ready agentic AI systems from code to deployment. This focused session presents a complete development stack through live demonstrations, technical deep dives, and interactive discussions.

What You Will Learn:
✔ Production-Ready Agentic AI – See live demonstrations of the complete development stack for building agentic AI systems, from coding environments through deployment and monitoring.
✔ Technical Implementation Strategies – Gain insights from AWS and partner experts on deploying agentic AI solutions with real-world use cases and AWS integration patterns.
✔ Expert Q&A & Networking – Engage directly with technical leaders from AWS and partner companies at dedicated booth stations, plus connect with fellow developers and practitioners building agentic AI applications.

Agenda

• 4:30PM - 5:30PM Check-In, Food and Booth Visiting
• 5:30PM - 7:45PM Kick-off, Tech talks, Live Demo
• 7:45PM - 8:30PM Networking, Q&A over Booths

Featured Speakers:

• David Potes, Sr. Manager, Data & AI at AWS
• Victor Wang, Staff Software Engineer at Deepgram
• John Wang, VP at Saviynt
• Kevin Locke, Sr. Director at 4Minds
• Sachin Mohan, Director at Zendesk

#####

Federated Intelligence in Production with Rhino CF and NVIDIA

Thu, Jul 9 · 5:30 PM PDT

135 Constitution Drive, Menlo Park, CA 94025, Menlo Park, CA, US

Welcome to the AI meetup in Silicon Valley, in collaboration with Rhino CF and NVIDIA. Join us for deep dive tech talks on Federated AI and Agent,

Agenda:
- 5:30pm~6:00pm: Checkin, food and networking
- 6:00pm~6:10pm: Welcome, Community update
- 6:10pm~8:00pm: Tech talks and Q&A
- 8:00pm~8:30pm: Open discussion, Mixer and Closing.

Talk 1: Operationalizing Federated Learning: Deploying, Scaling, and Integrating Models, Agents, and Apps.

Speaker: Adrish Sannyasi, Vice President, Customer Solutions and Delivery, Rhino Federated Computing Platform


Abstract: Federated learning enables collaborative model training across distributed data sources without centralizing sensitive data, making it an attractive approach for regulated and data-intensive industries. However, real-world deployments introduce challenges that go far beyond research prototypes, including heterogeneous infrastructure, data imbalance, data source connectivity, security, governance, and operational scalability.
This interactive lunch & learn focuses on federated learning as deployed in production systems, emphasizing industry-driven use cases such as drug discovery, federated clinical data networks, financial fraud detection, cybersecurity, and automotive systems, among others. Using the Rhino Federated Computing Platform (FCP) and NVIDIA accelerated infrastructure and NVIDIA FLARE framework software, the session examines how federated learning workflows are implemented across cloud and on-prem/edge environments.
Through concrete deployment examples and a guided user workflow and technical walkthrough, attendees will explore how federated learning jobs are orchestrated across heterogeneous multi-cloud systems, how aggregation and policy enforcement rules are applied, and how real-world issues such as security, compliance, and system scaling are handled. The session will also highlight lessons learned from moving federated learning systems from pilot phases into production in multiple industries, including infrastructure setup, FL data contribution cycle, data contribution incentive management, system performance optimization, and model and data lifecycle management.
Additionally, the session extends beyond model generation to demonstrate the complete production lifecycle. We will examine the transition from collaborative training to decentralized inference and decentralized agents, illustrating how a global model—such as a fraud detection system or a molecular property prediction model trained across multiple banks/pharma companies—is deployed back to local environments to score live data. We will then discuss the "last mile" of FL: integrating these inference signals into downstream applications.

Talk 2: Building Production Federated AI with NVIDIA FLARE: APIs, Operations, Agents, and Enterprise Scale

Speakers:
Chester Chen, Senior Product & Engineering Manager, NVIDIA Federated learning, FLARE


Peter Cnudde, Director of Engineering, NVIDIA Federated learning, FLARE


Abstract: This session explores how NVIDIA FLARE supports the full lifecycle of production federated AI, from application development to deployment, operations, automation, and future agent-assisted workflows. We will begin with the modern FLARE API direction, including the Client API for converting existing ML code into federated workloads, the Recipe API for reusable job construction, and the FLARE API/CLI for programmatic job submission, monitoring, and system operations.
We will highlight key FLARE 2.8.0 capabilities that make real-world deployments easier to operate: modern NVIDIA FLARE CLI automation with JSON/schema support, docker and Kubernetes deployment preparation, Docker/Kubernetes job launchers, multi-study support, live log streaming, tensor disk offload for large-model FedAvg, improved large-model streaming reliability, and new examples for multimodal, language-model, financial-services, and privacy-oriented workflows.
The session will also look ahead to where FLARE is heading: Collaborative API support for more flexible federated algorithms and analytics, FLARE agentic readiness with skills and benchmarks, Auto-FL research loops, Slurm/HPC support, memory-management enhancements, and integration with enterprise applications. Attendees will leave with a practical understanding of how FLARE is evolving into a production platform for secure, scalable, and increasingly automated federated AI across cloud, on-prem, edge, and HPC environments.

Venue:
SVAI Hub, 135 Constitution Drive, Menlo Park, CA 94025