Lakeside AI: Bringing Models to Data with Open Lakehouses & Agents
RAG, AI Infrastructure, Streaming, Open Data Lake
As AI workloads scale, data ingestion has become the silent bottleneck. Models demand high-quality, current, and compliant data—yet most pipelines remain brittle and opaque. This talk dives into how managed ingestion frameworks solve that challenge by integrating streaming and batch data into Iceberg tables with full governance.
We’ll explore ingestion automation patterns—schema inference, type evolution, deduplication, compaction, and CDC handling—alongside built-in observability and lineage capture. Using open technologies such as Apache Iceberg, Trino, and Starburst’s ingestion orchestration, you’ll see how to ingest once and serve many downstream AI use cases, from analytics to RAG.
Attendees will leave with a clear blueprint for turning ingestion from a maintenance burden into a competitive advantage for AI-driven organizations.
Key Takeaways:
• How governance, lineage, and compaction make ingestion AI-ready.
• Real-world architecture patterns for scalable managed ingestion.
Jitender Aswani is SVP of Engineering at Starburst, leading the company’s AI and data platform strategy. His teams build Starburst’s open, federated data infrastructure, powered by Trino and Apache Iceberg, to help enterprises unify analytics and AI across clouds, SaaS, and on-prem systems.
Mainak Ghosh is a software engineer at Starburst, working to help customers adopt open lakehouse for AI using managed ingestion and transformation. Previously, he was on the Kafka and SQL teams at Twitter, building large-scale real-time data platforms. Mainak holds a PhD from University of Illinois Urbana-Champaign, where I researched big data storage systems.
Martin Traverso is CTO at Starburst Data and Co-founder of the Trino Software Foundation. With a strong background in computer science and engineering, Martin has a track record of technical excellence in various organizations, including Facebook and Proofpoint. He holds both a Master's and Bachelor's degree in Computer Science from Drexel University.