5 Startups Making Big Data Business Friendly

These five Big Data analytics startups are all about making it easier for business users to find, access and analyze Big Data.

Using a Web crawler to scrape data from the Internet can produce useful business data such as pricing, reviews or contact data. But crawlers require technical expertise both to collect and clean the data.

Datafiniti makes scraping product, business and property data from the Internet easy for business users in two ways.

First, it uses a cloud-based Web crawler, 80legs, so users can log into a browser to launch custom queries. This runs on Amazon Web Services, which means it can be integrated into existing business application and will scale accordingly. Second, it cleans up the data using new technologies such as natural language processing and automated validation application so the data is immediately usable.

Big Boast: Nickelodeon used Datafiniti to ensure its branded products were priced correctly for families, no matter where they were sold online.

Finding the right data is a major challenge in enterprises with Big Data. Silos of data and the sheer volume of databases and systems make it nearly impossible for business users to know where key data is stored. Alation is one of a growing number of companies targeting this problem by allowing business users to search enterprise data.

Alation doesn’t just catalog the data. It also incorporates machine learning that tracks who and which department used which datasets. Alation allows end users to share input about the data through tagging, comments and rating. Finally, it narrows results by eliminating duplicate sets and schemas.

Big Boast: Teradata made Alation a strategic partner for data cataloging in its Universal Data Architecture.

One of the challenges with Hadoop datasets is that it takes some sort of IT intervention to make the data useful for business intelligence users. In the past, that meant integration work or moving key data into a separate OLAP data warehouse or data mart for use with BI tools.

Kyvos Insights solved this problem by developing OLAP software that works directly on Hadoop. Its visual, drag-and-drop interface eliminates the need for coding, so business users can create and analyze "cubes" directly on Hadoop. This ability to support OLAP-based dimensional analytics of Big Data on Hadoop is what makes Kyvos unique, according to TDWI Research Director Philip Russom.

Kyvos works on all major Hadoop distributions and supports Microsoft Excel and Tableau. Other solutions that bring OLAP cubes to Hadoop include AtScale and Teradata's Think Big Analytics, according to a 451 Research report published last year.

Big Boast: CRN Magazine named the flagship product, Kyvos, one of the coolest big data products of the year in its "Tech Midyear in Review."

Arcadia Data also uses the OLAP model for tapping into Hadoop data, but while other players try to connect existing BI with Hadoop, Arcadia wants to reinvent business intelligence for Hadoop. That’s a key difference, because instead of pulling from Hadoop, Arcadia Data uses Hadoop as an operating system and runs directly on Hadoop servers to access large datasets.

You can test-drive Arcadia Data by downloading Arcadia Instant, a free version of the company’s front-end data visualization tool. It’s available for Windows and Macintosh. The full enterprise version includes advanced analytics tools, plus other enterprise features such as security, scaling, visualization and optimization.

Big Boast: Arcadia Data offers unified front-end visual analytics tools and BI platform. It also recently launched a user community, including a forum and a library of articles and tutorials.

JethroData is another option for bringing Hadoop data into your BI solutions. While other software focuses on OLAP cubes, JethroData is a SQL-on-Hadoop solution. It acts as an "acceleration layer" by indexing select Hadoop datasets, rather than performing a full scan.

That data is then cached on a server between BI and the Hadoop datasets. BI queries are run against the stored indexes, rather than the full datasets. Since it uses a standard connector, JethroData can be used with your existing business intelligence tools. It’s also compatible with major Hadoop distributions such as Amazon, Cloudera, Hortonworks and MapR.

Big Boast: The company claims its engine produces results up to 100 times faster than alternative SQL-on-Hadoop solutions.