QCon AI New York 2026 Announces Initial Sessions, Focuses on AI Systems Engineering

QCon AI New York 2026 Program Details

QCon AI New York 2026 has confirmed 23 of its more than 30 planned sessions. The conference, scheduled for December 15–16, 2026, at The Westin Jersey City Newport, targets senior engineers, architects, and technical leaders involved in production AI systems. Initial talks will cover identity and authorization for autonomous agents, guardrails for operations agents on Kubernetes, shared model-serving platforms, and post-deployment evaluation of AI decision systems.

Shift to AI Systems Engineering

According to Conference Chair Hien Luu, the program reflects a significant shift in AI engineering towards systems engineering. The focus is moving from individual model behavior to the broader system behavior of AI. This includes managing bounded execution authority for agents, handling context and state, and integrating probabilistic models within deterministic control planes.

Core infrastructure now includes harness engineering, continuous evaluation, observability, and policy enforcement. Architectural constraints increasingly incorporate inference economics such as latency, token usage, model routing, and cost.

Interdisciplinary Approach

Luu noted that the evolving discipline of AI engineering is drawing from fields like distributed systems, security, platform engineering, and site reliability engineering. The goal is to ensure AI systems are reliable, observable, controllable, and economical to operate in production environments.

Keynote on Agent Identity and Authority

Nancy Wang, CTO at 1Password, will deliver a keynote titled "When Software Becomes a User: Identity and Authorization for Agents in Production." This session will explore how traditional identity and authorization models must adapt when software agents become active users of production systems.

The keynote will address challenges such as agents acting for multiple users, invoking unenumerated tools, creating subagents, and operating without direct supervision. Topics will include delegated authority across agent chains, auditability for multi-hop tool calls, and methods for granting agents necessary access without exposing underlying credentials.