Maana Announced – Big Data Search Engine Powered by Apache®Spark™

Print Friendly, PDF & Email

maana-logoMaana, a pioneer in search engine technology for big-data-fueled solutions, announced it has successfully built its industry agnostic, end-to-end search and discovery platform on Apache®Spark™. Spark enables Maana to perform massive-scale processing for machine learning and data mining. With Spark, Maana successfully overcomes the limitations of Hadoop’s MapReduce framework, especially with regard to performance and integration.

Maana is the first and only big data search engine powered by Spark. Switching to Spark was a business decision we didn’t take lightly, and afterwards we never looked back,” said Babur Ozden, founder and CEO of Maana.


Maana runs natively on Spark and uses its in-memory caching, which enables Maana to re-use parts of the computation. The in-memory caching is designed to increase application performance by holding frequently requested data in memory, reducing the need for database queries to get that data.

Maana has been running on Spark for over a year and a half. For Maana, working with extremely large technical datasets from numerous different sources requires both speed and sophisticated analysis. We weren’t getting the results with Hadoop MapReduce v1. Spark, the industry standard, is easier to use and the best alternative to the more exotic solutions available,” said Donald Thompson, founder and CTO of Maana. “The growing ecosystem around Spark allows our data scientists and our customers to use the languages and tools they are already familiar with.”


Sign up for the free insideBIGDATA newsletter.

Speak Your Mind