Vasudha Khare

Data today lives in two worlds: at rest, where you can query it, and in motion, streaming through systems like Apache Kafka®, built for throughput and ordering, not for querying. Bridging them means copying data out of the stream through connectors and stream processors into a warehouse or OLAP database, and every hop adds latency, a new copy, and one more thing that can break. This does not work for agents. Agents decide what to ask at runtime, chain dozens of retrievals per task, and act on what they read, so latency stacks up and stale data leads to wrong actions and consequences. Now lightning queries close that gap: a stream-native OLAP engine that queries streaming data directly, without ever copying it out. In this talk, you'll walk away with a working model for designing a stream-native query engine, one that holds fresh log records and historical columnar data in one place and answers queries consistently across both, with no gap between what just happened and what you can query. You'll see how lightning queries serve ad-hoc filters and lookups without pre-built cubes while staying low-latency, and the API design choices that made that possible. Most importantly, you’ll…

Fresh off the keynote? Join us for an interactive follow-up discussion and open Q&A on the Streamhouse framework. We'll dive deeper into how this open, vendor-neutral data architecture keeps real-time context continuously available for modern applications and AI agents.

AI systems don’t fail because models are weak. They fail because context is wrong. As agentic AI moves from experimentation to production, teams are discovering a new class of data problems. Context is scattered across operational databases, event streams, APIs, and vector stores. It’s stale by the time it reaches the model, inconsistent across tools, and expensive to recompute. Most architectures were never designed to continuously assemble and serve context at runtime. This talk introduces context engineering as a practical, systems-level discipline focused on solving these data challenges. Rather than treating context as static input, context engineering treats it as a continuously computed product, derived from live business signals, enriched in real time, governed, and served with low latency to AI systems. We’ll focus on the role of streaming and event-driven architectures as the foundation for this approach. You’ll see why batch pipelines and warehouse-centric designs struggle with agent workloads, and how stream processing enables data enrichment and reprocessing of context as data evolves. In the second half, we’ll build this live. Using Kafka and Flink, we’ll construct a…