
Frontier Signals #01: Infrastructure Behind Physical AI — Intelligence Beyond the Screen
Frontier Signals #01: infrastructure Behind Physical AI
Building the Infrastructure for Physical AI
AWS Builder Loft | August 10
Intelligence Beyond the Screen
Artificial intelligence is moving beyond the screen.
The next generation of AI will not simply generate text, images, and code. It will perceive its surroundings, reason about changing environments, simulate outcomes, make decisions, and take action in the physical world.
As AI expands into robotics, autonomous systems, manufacturing, logistics, mobility, industrial automation, and embodied agents, an entirely new infrastructure stack is emerging to support it.
Frontier Signals #01: Physical AI brings together founders, engineers, researchers, investors, and infrastructure leaders building the technologies that will power this next generation of intelligent systems.
The evening will feature rapid keynote talks followed by two technical panel discussions exploring the compute, cloud, data, agent, and developer infrastructure required to move AI from models to machines.
The central question of the evening is:
What infrastructure must exist between an AI model producing an output and an intelligent machine taking a safe, reliable action?
This is not a general introduction to AI. It is a focused technical gathering for people building the systems beneath Physical AI.
🎤 Frontier Keynotes
A series of rapid keynote talks highlighting the technologies shaping the future of Physical AI.
Topics include:
The transition from digital intelligence to physical intelligence
GPU compute and AI infrastructure
Cloud platforms for autonomous systems
Real-time inference and scalable AI deployment
AI agents and orchestration systems
Developer infrastructure for Physical AI
Investing in the next generation of AI infrastructure
🤖 Panel One
The Infrastructure Behind Physical AI
As AI systems move into the real world, infrastructure becomes the defining challenge.
This discussion explores the technologies enabling intelligent systems to operate reliably at scale.
Topics include:
Why Physical AI is the next major platform shift
Compute, inference, and GPU infrastructure
Cloud-native AI architecture
Real-time data infrastructure
World models and simulation
Agent infrastructure and orchestration
Edge versus cloud execution
Infrastructure investment opportunities
🔁 Panel Two
Building AI Systems for the Physical World
Building Physical AI requires more than powerful models—it requires systems that can continuously observe, reason, act, evaluate, and improve in dynamic environments.
This panel explores how intelligent systems move from demonstrations to reliable production deployments.
Topics include:
From foundation models to autonomous systems
Robotics, manufacturing, logistics, and industrial AI
Multi-agent systems
Loop engineering for autonomous execution
Reliability, evaluation, and observability
Memory and long-running AI systems
Failure detection and recovery
Developer tools and infrastructure for AI builders
🗓️ Agenda
5:00 PM – 5:30 PM
🍽️ Check-In, Food & Networking
Meet founders, engineers, researchers, investors, and AI infrastructure leaders while enjoying food before the program begins.
5:30 PM – 5:40 PM
Opening Remarks
The Next Wave of AI: From Foundation Models to Physical AI
An overview of how AI is evolving beyond digital assistants into real-world intelligent systems—and why infrastructure will define the next decade of innovation.
5:40 PM – 6:10 PM
🎤 Frontier Keynotes
A series of rapid keynote talks covering the most important technological shifts shaping Physical AI.
6:10 PM – 6:15 PM
Panel Transition
Connecting the keynote themes to the first technical discussion.
6:15 PM – 6:45 PM
🤖 Panel One
The Infrastructure Behind Physical AI
A discussion covering compute, GPUs, cloud infrastructure, data systems, AI agents, simulation, and the infrastructure required for real-time intelligence.
6:45 PM – 7:00 PM
💬 Audience Q&A
Open discussion with the first panel.
7:00 PM – 7:15 PM
🍽️ Networking Break
Continue conversations over food and meet fellow builders.
7:15 PM – 7:45 PM
🔁 Panel Two
Building AI Systems for the Physical World
A discussion on deploying AI into robotics and autonomous systems, covering loop engineering, multi-agent architectures, reliability, observability, evaluation, and production deployment.
7:45 PM – 8:00 PM
💬 Audience Q&A
Open discussion with the second panel.
8:00 PM – 8:20 PM
🤝 Structured Networking
Connect with founders, engineers, researchers, investors, and operators building the next generation of AI infrastructure.
8:20 PM – 8:30 PM
Closing Signals
Key takeaways, emerging opportunities, and the open problems that will define the future of Physical AI.
What You’ll Learn
Participants will leave with a clearer understanding of:
Why Physical AI represents the next frontier of artificial intelligence
The infrastructure required to deploy AI beyond software
Compute, GPU, cloud, and data systems for real-time intelligence
The role of AI agents, orchestration, and developer infrastructure
How world models and simulation enable autonomous systems
Reliability, evaluation, and observability for production AI
Where the biggest opportunities exist for founders, engineers, and investors
Who Should Attend
AI & Machine Learning Engineers
Robotics & Autonomous Systems Builders
Infrastructure & Systems Engineers
GPU & High-Performance Computing Engineers
Cloud & Platform Engineers
AI Agent Developers
Startup Founders
Researchers
Product Leaders
Investors exploring AI infrastructure and Physical AI
Students and early-career builders interested in frontier AI
About Frontier Signals
Frontier Signals is a recurring keynote and panel series exploring technologies before they become mainstream.
Each edition brings together the founders, researchers, engineers, investors, and technology leaders building the next generation of computing, artificial intelligence, and emerging infrastructure.
Important Registration & Entry Information
⚠️ Please read before registering:
Only registrations completed through the official AWS Builder Loft registration page will be accepted at the door. A Luma RSVP alone does not guarantee entry.
This event is strictly 18+.
All attendees must present a valid, government-issued physical photo ID. Digital copies or photos of IDs will not be accepted.
Personal vehicles are not permitted on the premises.
The AWS Builder Loft does not provide parking or storage for cars, bicycles, scooters, or similar personal transportation devices.
The official AWS registration link will be added once available.
🍽️ Food will be provided. No alcohol will be served.
AWS Builder Loft, 525 Market St, Courtyard Entrance, 525 Market St, San Francisco, CA 94105, USA
VeibeskrivelseSkann med kameraet – arrangementet åpnes i Somo-appen.









