in Case Studies


Top 5 SingleStore Showcase Applications and Product Demonstrations

Kevin White

Kevin is on the marketing team and is responsible for marketing operations and content strategy.

Top 5 SingleStore Showcase Applications and Product Demonstrations

We live in an age of digital realities. Most “goods” on the market are not physical entities, but more often software. However, consumers still need a tangible way to evaluate these digital products and services they are purchasing.

Enter the product demonstration – the best way to showcase the value of software in a short time.

At SingleStore, we pride ourselves on showcasing relatable product demonstrations from the main stage at Strata+Hadoop World to our locally hosted meetups.

Here’s a look at our top five showcase applications over the past year:

power-stream-io-t-applicationPowerStream IoT Application

PowerStream is an Internet of Things (IoT) application that displays visualizations and alerts from approximately 20,000 wind farms and 200,000 individual wind turbines. Temperature and vibration input from millions of sensors on turbines is ingested through Apache Kafka and inserted into SingleStore for real-time data exploration. Using this data and a machine learning algorithm, PowerStream predicts and visualizes the health of each turbine up to the last second.

Dataset
2 million sensors across 197,000 global wind turbines.

Results
Less than $20k annual cost. Runs on seven c4.2xlarge AWS instances at approximately $0.311 per instance/hour.
Processing more than 1 million transactions per second.


real-time-analytics-at-pinterestReal-Time Analytics at Pinterest

Pinterest uses SingleStore Streamliner to ingest data into SingleStore using Apache Kafka and Spark Streaming. In this application, Pinterest is filtering trending Pins across the United States and enriching data by adding geolocation and Pin category information. As data streams in, Pinterest is able to run queries in SingleStore to generate engagement metrics and report on various event data like pins, repins, comments and logins.

Read the full Pinterest Engineering Blog Post here: https://engineering.pinterest.com/blog/real-time-analytics-pinterest

Dataset
Pinterest Pin data across the United States.

Results
After integration, Pinterest now has a system that allows analysts to use SQL to explore data and derive insights in real time.


supercar-real-time-geospatial-intelligenceSupercar: Real-Time Geospatial Intelligence

Supercar leverages SingleStore geospatial capabilities to track and analyze 170 million real world taxi rides. By sampling this dataset and creating real-time records while simultaneously querying data, Supercar simulates the ability to achieve real-time insights across hundreds of thousands of objects on the go.

Dataset
170 million real world taxi rides.

Results
Query and visualize geospatial data with SQL.
Insights hundreds of thousands of objects in under a second.


taxistats-real-time-geospatial-dashboardTaxistats Real-Time Geospatial Dashboard

Taxistats is a showcase application that makes use of the same dataset used in Supercar. Data is streamed into SingleStore and visualized with Zoomdata to simulate pickup and drop-off information from real taxi ride data. The Taxistats dashboard displays data as it is collected while simultaneously running SQL analytics.

Dataset
170 million real world taxi rides.

Results
Sub second SQL queries on millions of taxi rides.


Our Wikistats application compares the popularity of entries from Wikipedia’s public dataset consisting of more than 250 billion records. Similar to Google Trends, multiple search terms can be queried and trends are visualized over the course of a year. When a query is run, SingleStore scans over 30 billion records and returns results in less than 200 milliseconds.

Dataset
Wikipedia public dataset of 250 billion records.
30 billions rows scanned in under 200 milliseconds.

Results
10x faster and 60x more cost effective than Oracle 12c.
Runs on a 6-node cluster that costs about $5/hr on Amazon or less than $20k.
Scans and aggregates tens of billions of records in milliseconds.




Share