Nihar Sheth

Some Lambda workloads run constantly. For steady-state, high-throughput functions, per-invocation pricing becomes more expensive than running the same code on EC2, but you don't want to manage instances.

Building Kafka consumers today means running, scaling, and patching compute infrastructure yourself, and writing your own retry, poison-pill, and observability plumbing. AWS Lambda's event source mapping for Apache Kafka removes that operational burden, letting you process streams from AWS MSK, self-managed Kafka, and Confluent Cloud with nothing more than function code. In this session, see how to build low-latency, high-throughput consumers that scale automatically with partition lag, decode and validate records against Confluent Schema Registry, and cut cost by filtering JSON, Avro, and Protobuf events before invocation. The same event-driven pattern can trigger AI agents directly from your streams, so events can kick off agentic workflows without extra glue code. We cover tuning for throughput and latency, extensive CloudWatch metrics and logs for consumer visibility, and error handling patterns including retries, bisect-on-error, and on-failure destinations for dead-letter queues. By attending this session, your contact information may be shared with the sponsor for relevant follow up for this event only.

Gain a deep understanding of how AWS Lambda uses event source mapping to retrieve items from streams and queues. Explore the inner workings of event source mapping and how its behavior differs across seven different streaming and queuing services, such as Amazon Kinesis and Amazon SQS.

With business-critical analytics and decision-making processes relying increasingly on data warehousing workloads, batch processing is becoming a thing of the past and organizations are investing more into near real-time data ingestion along with advanced analytics and machine learning jobs to make decisions and grow…

Amazon Kinesis makes it easy to collect, process, and analyze real-time streaming data so that you can get timely insights and react quickly to new information. In this chalk talk, learn from best practices for building serverless streaming architectures to ingest and process billions of events per day at low latency.