Hortonworks Inc. is extending its sphere of influence deeper into the Hadoop ecosystem with a sweeping plan to support the Spark in-memory processing framework. The move is the latest in a string of initiatives from the software maker that are designed to make Hadoop more attractive for enterprises.

Hortonworks  hopes that the initiative will give it a leadership role in the development of what is one of the most strategically important sub-projects in the Hadoop ecosystem. It will have a battle on its hands, however. Cloudera, Inc. announced support for Spark early this year, as did MapR through an alliance with Databricks, Inc., whose founders developed Spark.

Born out of the celebrated AMPLab at UC Berkeley, Spark offers a faster and simpler alternative to the default MapReduce computational engine in the batch processing framework. With its easy-to-use interface, the software lets users crunch data 10 times faster when running on disk and up to 100 times quicker when leveraging memory, a tremendous edge that has drawn a lot of community interest.

But while it may have cemented its position as the top contender to MapReduce’s throne, Spark isn’t ready for prime-time quite yet, lacking many of the features needed to support large-scale analytical environments. The new plan from Hortonworks aims to hurry along the project’s maturation and align it with the needs of its enterprise customers.

The company has seemingly modeled the roadmap after its freshly completed Stinger Initiative to modernize the Hive data warehouse for Hadoop, drawing on the lessons from that initial success and devoting the first leg of the journey to laying down a groundwork for the main enhancements.

Hortonworks will start off by making Spark compatible with a number of other strategically important technologies, including Hive, the Ambari management engine (whose development it’s sperheading as well) and popular user directories. The latter addition paves the wave for automated authentication, which the company said is one of the most common feature requests it receives from enterprise buyers.

The second phase of the project will see the firm extend that integration to the nerve center of Hadoop, YARN, in an effort to optimize hardware utilization. Hortonworks intends to double down on the security aspect of the initiative, too, through support for Project Argus, the open-source version of the technology it obtained through the acquisition of XA Secure Inc. earlier this year. Topping off the upgrade is integration with the YARN Application Timeline Server (ATS), which provides a way to retrieve operational data about applications that can be useful for monitoring and maintenance.

The initiative is set to complete in early 2015, with the first batch of features currently in preview.