Powering Real-Time Analytics with Apache Kafka and Rockset | Streaming Audio: Apache Kafka® & Real-Time Data podcast

‌

Episode description

‌

Episode description

Using large amounts of streaming data increasingly requires interactive, real-time analytics and dashboards—and this applies to any industry, including tech. CTO and Co-Founder of Rockset Dhruba Borthakur shares how his company uses Apache Kafka® to perform complex joins, search, and aggregations on streaming data with low latencies. The Kafka database integrations allow his team to make a cloud-native analytics database that is a fundamental piece of enterprise infrastructure.

Especially in e-commerce, logistics and manufacturing apps are typically receiving over 20 million events a day. As those events roll in, it is even more critical for real-time indexing to be queried with low latencies. This way, you can build high-performing and scalable dashboards that allow your organization to use clickstream and behavioral data to inform decisions and responses to consumer behavior. Typically, the data follow these steps:

Events come in from mobile or web apps, such as clickstream or IoT data

The app data is sent to the cloud

Data is fed into the database in real time

This information is shared live on a dashboard or via SaaS application embeds

For example, when working with real-time analytics in real-time databases, both need to be continuously synced for optimal performance. If the latency is too significant, there can be a missed opportunity to interact with customers on their platform. You may want to write queries that join streaming data across transactional data or historical data lakes, even for complex analytics. You always want to make sure that the database performs at a speed and scale appropriate for customers to have a seamless experience.

Using Rockset, you can write ANSI SQL on semi-structured and schemaless data. This way, you can achieve those complex joins with low latencies. Further data is required to supplement streaming data, but it can be easily supported through supported integrations. By having a solution for database requirements that are easily integrated and provide the correct data, you can make better decisions and maximize the result.

EPISODE LINKS