Apache Hudi Meetup @ Onehouse Bengaluru Office

Join us for an Apache Hudi Open Source Meetup, hosted by Onehouse at COWRKS Ecoworld, Bangalore! This edition focuses entirely on the next-generation Apache Hudi 1.2 architecture — exploring how its major structural overhauls power modern AI/ML workloads. Ideal for data engineers, lakehouse architects, and AI/ML practitioners.

👉 Registration: Open — RSVP to secure your spot: https://forms.gle/UboRNekS8FQjcZjCA
📅 Date & Time: Wednesday, July 29, 2026 · 4:00 – 7:00 PM IST
📍 Venue: COWRKS Ecoworld, 10th Floor, Building 4D, ECOWORLD, Outer Ring Rd, Devarabisanahalli, Bellandur, Bengaluru, Karnataka 560103 —[ https://maps.app.goo.gl/Q7tB1M6bYCtQBU529](https://maps.app.goo.gl/Q7tB1M6bYCtQBU529)

⏰ Agenda

• 4:00 – 4:30 | Welcome & Introductions — kickoff, opening remarks, community announcements
• 4:30 – 6:45 | Hudi Tech Talk Sessions — titles & speakers TBA
• 6:45 – 7:00 | Open Q&A, Networking & Wrap-up

💡 What you'll learn

• Multimodal Data as a First-Class Citizen: How Hudi 1.2.0's new native logical types — VECTOR, VARIANT, and BLOB — plus the read_blob() and hudi_vector_search() Spark SQL functions bring embeddings, semi-structured payloads, and binary objects (images, audio) into the lakehouse alongside structured data, backed by RFC-99/100/102.
• Optimized Storage for AI/ML workloads: Native Lance columnar format integration for AI/ML access patterns, and how Hudi's indexing primitives (record-level, secondary, expression indexes) now extend to accelerate vector search and point lookups — no separate vector DB/document store required.
• Streaming at Scale: Full Record Level Index (RLI) support landing on Flink — global indexing, efficient record routing, and dynamic bucket scaling — plus the new FLIP-27 Source V2 with resumable split assignment and stronger pushdown capabilities.

👥 Who should attend

• Data, Platform, and Cloud Engineers building real-time lakehouses.
• ML/AI engineers building high-throughput feature stores, RAG architectures, and vector-search layers.
• Engineers working with Apache Spark, Flink, Trino, and cloud-native computing fabrics.
• Open-source contributors interested in advanced transactional table formats.

📝 RSVP: Entry is free, but seats are limited by venue capacity — please register in advance so we can pre-generate your visitor pass at the tech park security desk.

See you there — let's build the AI-native lakehouse together!