Bangalore Iceberg Community Meetup

Note: Meetup RSVPs are closed. Please register on Luma directly.

Apache Iceberg has become the backbone of the modern open lakehouse, and Bangalore's data community is diving in. Join ClickHouse, e6data and OLake for a morning dedicated to all things Iceberg: open table formats, real-time analytics on the lakehouse, and the query engines that bring it all together.
Whether you're already running Iceberg in production, evaluating a migration from a traditional warehouse, or just curious about what the lakehouse hype is really about, this meetup is for you.

Don't miss out! RSVP and secure your spot!

🗓️ Agenda:

• 09:30 AM: Registration & networking
• 10:30 AM: Welcome & opening
• 10:40 AM: Talk 1 - ClickHouse and the Data Lake: Under the Hood by Shankar Iyer, Senior Software Engineer @ ClickHouse
• 11:10 PM: Talk 2 - Iceberg V2 or V3: What Should You Run in Production? by Rajath Gowda (Founding Engineer, e6data) and Lakshmi Narayana G (Founding Engineer, e6data)
• 11:30 PM: Break
• 11:45 PM: Talk 3 - Apache Iceberg in Production: The Adoption Bottlenecks by Vaibhav Verma, Backend Engineer @ OLake and Ankit Sharma, Lead Engineer @ OLake
• 12:15 PM: Closing Remarks
• 12:30 PM: Lunch & networking

If anyone from the community is interested in sharing a talk at this or future events, complete this CFP form and we’ll be in touch.

🎤 Session: ClickHouse and the Data Lake: Under the Hood
Description: Data lakes are evolving from collections of Parquet files into managed, interoperable table formats such as Apache Iceberg and Delta Lake. This session explores how ClickHouse is evolving alongside them—from querying files to understanding metadata, catalogs, snapshots, deletes, writes, and table maintenance.
We’ll look at the engineering work behind this evolution, including file and metadata pruning, caching, distributed execution, open-table-format support, and Data Lake catalogs. The session will connect these features to the broader architecture of ClickHouse as a high-performance compute engine over modern data lakes.
Speaker: Shankar Iyer, Senior Software Engineer @ ClickHouse

🎤 Session: Iceberg V2 or V3: What Should You Run in Production?
Description: The talk asks one question: does your Iceberg table actually need V3, or is V2 still fine? It covers how V2 handles row-level change and where deletes get expensive, files pile up, scans crawl, compaction becomes a standing job. Then it gets into what deletion vectors fix (one bitmap per data file instead of a pile of delete files) and what they don't, like small files and equality-delete overhead. Row lineage and Variant get the same honest treatment: genuinely useful for CDC and incremental reads, not worth much outside that. It also names the real upgrade costs: more work on writers, patchy engine support, and no quiet way back once you flip the switch. Three things stick with you by the end. V2 is still right for batch and append-heavy tables. V3 only pays off on merge-on-read and incremental workloads. And the real blocker is rarely the format version itself, it's whether every reader touching your table can handle V3 yet. A checklist closes the talk, meant for running against your own table before you commit.
Speaker: Rajath Gowda (Founding Engineer, e6data) and Lakshmi Narayana G (Founding Engineer, e6data)
Rajath Bio: Rajath Gowda is a founding engineer at e6data, where he works on data engineering and the systems behind high-performance lakehouse analytics. His work spans areas such as query performance, data lake architecture, and large-scale analytical workloads. Rajath also regularly shares his experience through technical writing and industry sessions on modern data infrastructure.
Lakshmi bio: Lakshmi Narayana is a founding engineer at e6data, where he works on high-performance data infrastructure, distributed systems, and query processing. A computer science graduate from IIIT Guwahati, he began his career in full-stack engineering before moving deeper into systems and performance engineering. At e6data, he focuses on building fast, scalable technology for modern analytical and lakehouse workloads.

🎤 Session: Apache Iceberg in Production: The Adoption Bottlenecks
Description: Iceberg is a solved problem on paper. ACID. Time travel. Row level changes. And interoperability, the feature it's best known for, which turns out to be mostly a paper feature once you run it across real engines. We'll cover what it takes to actually get there, along with the best practices for ingestion pipelines into Iceberg, exactly-once semantics, and what it takes to make queries fast.
Speaker: Vaibhav Verma, Backend Engineer @ OLake and Ankit Sharma, Lead Engineer @ OLake
Vaibhav Verma in building fastest ingestion pipelines to Apache Iceberg
Ankit Sharma is Apache Iceberg Enthusiast, Lead development of OLake and Fusion

What to expect:
Talks from engineers at ClickHouse, e6data, and the local community covering topics like querying Iceberg tables at speed, CDC pipelines into open table formats, catalog interoperability, and lessons from running lakehouse architectures in production. Expect deep technical content, live demos, and honest conversations about what works (and what doesn't).
We'll wrap up with networking over lunch, so stick around to swap notes with fellow data engineers, platform teams, and lakehouse enthusiasts from across Bengaluru.

Bring your questions, your war stories, and your appetite for fast queries. See you there!