Operational AI: Real-Time Analytics in Production

This meetup, hosted by the Open Source Analytics Community in Bangalore, focuses on Operational AI: Real-Time Analytics in Production. It is designed for engineers, data practitioners, and open-source enthusiasts interested in building AI-powered applications that react quickly while maintaining trust and accuracy. Attendees will gain practical insights into processing, analyzing, and acting on data at low latency, exploring how real-time analytics frameworks, streaming platforms, and open tooling power real-time intelligence and AI workflows. The session will feature practical talks, lessons learned from real use cases, and discussions on open technologies for ingesting, processing, and serving data with low latency for AI-driven applications.

Speakers and Topics:

• Josh Lee (Altinity): Building a Foundation for AI with ClickHouse® and Apache Iceberg Storage, focusing on adapting ClickHouse® for compute/storage separation on Iceberg table data to support AI use cases.
• Ayush Sawant (Nutanix): AI Monitoring 101: Why Your Old Dashboard Won't Work for LLMs, detailing a full-stack observability pipeline for scaling LLMs and Agentic workflows, covering infrastructure metrics and LLM-specific telemetry.
• Sasi Teja (Kafka & GlassFlow): When Events Become Decisions: Real-Time AI with Kafka, GlassFlow & ClickHouse®, demonstrating a real-time AI analytics pipeline for Customer 360 use cases, including anomaly detection and LLM analysis.
• Debabrata Panigrahi (Parseable): Debugging AI Agents in Production, addressing the observability gap in agentic AI and demonstrating how to auto-instrument agents with OpenTelemetry to pinpoint and fix issues.

Food and drinks will be provided, courtesy of Nutanix sponsoring the venue. This event offers actionable ideas for anyone working on real-time projects involving AI.