Skip to content

Details

Hello everyone! Join us for an IN PERSON Apache Kafka® meetup on Nov 28th from 11:00 am, hosted by Data-Axle!

📍 Venue:
Data-Axle, ICC Tech Park, Tower A – 8th Floor, Senapati Bapat Rd, Laxmi Society, Model Colony, Shivajinagar, Pune – 411016

Landmark: Opp. JW Marriott.

Maps Location: https://maps.app.goo.gl/HcRGNLDwviNRZ4ss9

***
Agenda:

  • 11:00- 11:10: Welcome & Introduction
  • 11:10- 11:50: Siddesh Vyavahare, Advisory Solutions Engineer, Confluent
  • 11:50- 12:30: Jyoti Gehlot, Consulting Architect, Elastic & Krishnan Ragavan, Founder, Kaizen Studio
  • 12:30 - 12:45: Break
  • 12:45 - 13:25: TBC
  • 13:25- 14:30: Lunch

If you would like to speak at or host a meetup please let us know! community@confluent.io (http://confluent.io/)

***
💡 Speaker:
Siddesh Vyavahare, Advisory Solutions Engineer, Confluent

Talk:
Real-Time Context for AI: Kafka and the Streamhouse

Abstract:
Kafka made streaming the backbone of modern data platforms, but analytics and AI still wait on batch pipelines. Streamhouse makes the stream the source of truth, so live data is fresh and queryable everywhere. Learn what it is, how it builds on Kafka, and where it fits in your architecture.

-----
💡 Speaker:
Jyoti Gehlot, Consulting Architect, Elastic & Krishnan Ragavan, Founder, Kaizen Studio

Talk:
Tools or Context? AI Agents on a Live Kafka Stream

Abstract:
Give AI agents the same live data and the same business questions, and you won't get the same answers. Jyoti Gehlot and Krishnan Ragavan built a social-pulse setup where shop events, Bluesky posts, and business facts flow through Kafka and Logstash into Elasticsearch. They then compared Agent Builder agents side by side: one using platform tools, one using custom ES|QL tools, and one backed by a Context Engine AI index.

On facts that lived only in the AI index, the context-backed agent answered 8 out of 8 correctly. The others answered none. Context isn't always the better choice, though: well-written ES|QL tools matched its accuracy at a lower cost. Jyoti and Krishnan also compare how fresh your data is when you push it through Kafka versus pulling it with a managed automation.

You'll see real token costs and the trade-offs they ran into, and you'll leave with a simple way to tell whether your agent needs better tools or better context.

-----
💡 Speaker:
TBC

Talk:
TBC

Abstract:
TBC

***
If you would like to speak or host our next event please let us know! community@confluent.io (http://confluent.io/)

Related topics

Events in Pune, IN
Apache Kafka
Big Data
Stream Processing
Open Source
Technology

You may also like