IN PERSON! Apache Kafka® Meetup Bangalore- May 2025

Hello everyone! Join us for an IN PERSON Apache Kafka® meetup on May 17th from 11:00AM, hosted by Traceable AI in Bangalore!

📍 Venue:
TRACEABLEAI SOFTWARE INDIA PRIVATE LIMITED
Spaces, 7th Floor, Fairway Business Park
Behind Dell Road, Next to Embassy Golf Links
Domlur, Bangalore, Karnataka, India – 560071
Google Maps link: https://maps.app.goo.gl/6es4YFtWpzpsLgqt9

***
Agenda:

• 11:00 - 11:10: Welcome
• 11:10 - 11:50: Laxman Ch / Ronak Kothari (TraceableAI), Engineering Architect
• 11:50 - 12:30: Mohtasham Sayeed Mohiuddin, Cloud Enablement Engineer - Kafka, Flink, AI, Confluent
• 12:30 - 12:40: Break
• 12:40 - 13:20: Shruti Mantri, Staff Engineer
• 13:20 - 14:30: Lunch

***
💡 Speaker:
Laxman Ch / Ronak Kothari (TraceableAI), Engineering Architect

Talk:
Scaling Kafka Streams in Production: Lessons from 6+ Years of Real-World Usage

Abstract:
In this session, we'll share practical insights from over six years of running Kafka Streams at production scale. We'll walk through the rationale behind choosing Kafka Streams, the scale requirements we operate under, and the architectural patterns we've evolved to meet them. Key topics include handling external API calls within stream processing, tenant isolation for bursty workloads, designing custom throttled punctuators, challenges with RocksDB usage, and building unified monitoring dashboards. Expect a candid walkthrough of what worked, what didn't, and the lessons we learned along the way. We also invite you to share your own experiences and approaches.

-----
💡 Speaker:
Mohtasham Sayeed Mohiuddin, Cloud Enablement Engineer - Kafka, Flink, AI, Confluent

Talk:
Detecting Fraud in Motion: Real-Time Fraud Detection with Flink, Kafka & AI

Abstract:
In this workshop, we'll explore how to harness the power of streaming data to detect fraud in real time using Apache Flink, Kafka, and AI on Confluent Cloud. You'll learn how to ingest transactional data, enrich it with customer and credit card streams, and use Flink to transform and prepare features for fraud detection models. We'll show how to integrate with AWS Bedrock for AI-generated insights and output fraudulent predictions—all in a scalable, cloud-native pipeline. Whether you're an engineer or a data scientist, this hands-on session will equip you to build intelligent, responsive systems that act on data the moment it arrives.

-----
💡 Speaker:
Shruti Mantri, Staff Engineer

Talk:
Handling Workflow Logs in Real-time using Kafka and Pinot

Abstract:
In complex environments, workflows are often long-running processes composed of multiple tasks. These workflows may extend from several minutes to several days, frequently pausing for manual actions, triggers, or external events. A Workflow Log Explorer typically provides an overview of different workflow executions, allowing users to drill down into detailed logs for each task within an execution.

In this talk, we will examine how Apache Pinot can be harnessed to achieve this complex, real-time collation of logs, enhancing visibility and traceability across lengthy and intricate workflows.