The Data Streaming Summit returns to San Francisco in 2026 — and this year, we’re bringing the community back to where it all began: Hotel Nikko.

As AI agents rapidly move from demos to production systems, one thing has become clear: agents require infrastructure built for real-time decision-making, durable memory, continuous context, event-driven orchestration, and operational reliability at scale. These are not entirely new challenges — they are problems the data streaming community has been solving for over a decade.

Data Streaming Summit 2026: The Data Streaming + Agent Infra Conference brings together the engineers, architects, founders, and operators building the next generation of AI systems alongside the streaming practitioners who power modern real-time infrastructure.

From Apache Pulsar and Kafka to agent runtimes, observability, governance, and orchestration frameworks, this is the event focused on building the infrastructure layer that enables intelligent, autonomous systems to operate reliably in production.

Attendees will explore how streaming technologies are shaping the future of AI agents through technical sessions, architecture deep dives, production case studies, and hands-on operational lessons from teams working at the forefront of real-time systems.

This year’s conference will focus on three core tracks:

Data Streaming Engines

Explore the technologies and architectures powering modern real-time systems. Topics may include Apache Pulsar, Kafka, Flink, Spark, stream processing, event-driven systems, multi-region infrastructure, scalability, performance optimization, reliability, storage, and real-time data platforms operating at massive scale.

Data Streaming for Agents

Learn how streaming infrastructure is enabling AI agents with real-time context, memory, retrieval pipelines, event-driven workflows, and continuous reasoning systems. Sessions may cover agent communication patterns, context pipelines, RAG architectures, streaming vector updates, real-time personalization, and production AI workflows.

Agent Harness, Runtime, and Governance

Dive into the emerging operational layer for AI agents. Topics may include agent runtimes, orchestration frameworks, observability, evaluation systems, governance, security, reliability, tool orchestration, human-in-the-loop systems, and infrastructure for safely operating autonomous systems in production environments.

Whether you’re building petabyte-scale streaming infrastructure, designing autonomous agent systems, or operating real-time AI applications in production, Data Streaming Summit 2026 is where the streaming and AI infrastructure communities come together to share ideas, solve hard problems, and shape the future of intelligent systems.