Surupa and Peter will share the latest progress from Meta's mission to build the future of human connection and the technology that makes it possible, including infrastructure for AI by leveraging agentic innovation across the stack.

Engineering at Meta is in the midst of a productivity revolution, driven by AI. Our goal is to dramatically increase engineering velocity by using agents to build new systems and operate existing ones. Teams are not just using AI to write code faster; they’re also leveraging AI to debug GPU failures, remediate system outages and more.

To unleash agentic productivity, we have embarked on a journey to safely enable agentic access to Meta’s systems. This talk will dive into what we’ve learned so far - building internal agents, deploying 3rd party agents, observing agents' interactions with our Infra, and deploying guardrails to prevent destructive actions.

To handle the scale and velocity of AI-written code, we will have no choice but to let AI manage our infrastructure too. Yet Andrej Karpathy recently described getting an app running in production as “assembling IKEA furniture”: cloud consoles, API keys, copy-pasted config, glue, ... all things that sit outside the code an LLM can reason about. Frontier models are trained on billions of lines of real languages like Python, TypeScript, and Go, and vanishingly little bespoke DSLs and manual procedures. By modeling infrastructure in code space so the LLM can do what it does best — code — we just need an oracle that can map code changes back to infrastructure outcomes. In this talk, I’ll share what we’ve learned at Pulumi working alongside leading AI companies and frontier labs to build for a world where agents manage infrastructure. The platforms that win in this new era will look different, but many of the human-ergonomic benefits of programming languages are what will get us there.

What if your best oncall engineer never slept, investigated every alert in parallel, and got smarter after every SEV? At Meta, we're building that reality. This talk covers how Production Engineering is deploying AI agents that autonomously investigate, diagnose, and increasingly mitigate production incidents across systems serving billions of users.
We'll deep-dive into RecInvestigator — an AI agent that has investigated 300+ SEVs across Meta's Recommendation Systems, cutting p50 MTDM by 51%. We'll explain how it encodes expert investigation workflows, correlates signals across Service Router, Tupperware, ODS, Scuba, and Presto, and maintains an ~80% match rate with human investigators. We'll then show how we're using AI-driven SEV pattern analysis to move from reactive firefighting to proactive prevention — identifying systemic failure modes, proposing architectural fixes, and prototyping agents that auto-mitigate service regressions during business hours.
The talk will cover the full journey we're on: (1) AI-powered pattern identification across hundreds of SEVs, (2) autonomous investigation agents that match human accuracy, (3) fixer agents that propose and implement reliability improvements in code, and (4) the moonshot — AI-native reliability where systems self-heal and development is reliability-aware by default. We'll share what worked, what didn't, and what we've learned about the trust boundary between autonomous agents and human engineers in production.

Panel information to be announced!

Surupa and Peter will share the latest progress from Meta's mission to build the future of human connection and the technology that makes it possible, including infrastructure for AI by leveraging agentic innovation across the stack.