BT

Facilitating the Spread of Knowledge and Innovation in Professional Software Development

Write for InfoQ

Topics

Choose your language

InfoQ Homepage News Adatao Launches Full Stack Data Intelligence Platform

Adatao Launches Full Stack Data Intelligence Platform

Adatao recently announced the general availability of its Data Intelligence platform. Its platform aims to make data analysis and predictive analytics available to everyone in large organizations. Adatao had secured an investment of $13 million last year from a group of investors including Bloomberg Beta, Lightspeed Venture Partners and Andreessen Horowitz.

The platform is comprised of three layers. At the foundation of the system lies the Big Data layer. Using Hadoop, Redshift, Amazon S3, RDBMS or local file sources, users can ingest data from heterogeneous data sources.

On top of it, there is the Big Compute layer, powered by an array of open source technologies like Apache Spark, Facebook’s Presto, Cloudera Impala and Apache Flink.. There lies the PredictiveEngine functionality, used by the user facing layer applications.

At the user facing layer, there are Narratives, Dashboards and Big R applications. Narratives is an environment that looks like Google Docs with extra features. Inside a document, users can insert content in text, images but also Python, R and SQL using a built in interpreter. Code inside the document can reference the underlying dataset and query it, generating up to date graphs. Using natural language, anyone in an organization can apply machine learning intelligence techniques like multilinear regression or k-means without being a data scientist himself.

Dashboards uses the same functionality and focuses on creating interactive dashboards by business users. Big R application focuses more on the data scientist needs of data cleansing and using R, Python, SQL or Scala to process data.

The system is using a proprietary architecture that is geared towards keeping most data in memory. This results in couple of seconds query time for a dataset of 120GB, without any pre-calculated aggregations. Lending from the heavy engineering background of the founders, Adatao plans to offer API integration in every layer of the platform. The Appbuilder can be used to build other applications like Narratives and Dashboards. The Appbuilder has both a visual interface designer for non-developers, several pre-defined application templates and also an SDK for developers to quickly develop applications for verticals such as Telecom, Financial Services and Retail. Adatao PredictiveEngine, Narratives and Dashboards are generally available, whereas Adatao BigR and AppBuilder are private beta for select customers and partners at the moment.

 

Rate this Article

Adoption
Style

BT