Ellicium: Benefitting from Analysis of Vast Amount of Unstructured and Machine Generated Data

Kuldeep Deshpande, CEO
“Data generated during human interactions can provide us wealth of knowledge. Similarly we need to tap into ability of machines to provide data and gain insights,” states Kuldeep Deshpande, CEO, Ellicium Solutions. “To help organizations leverage business efficiency by extracting wealth of information from unstructured data, we deliver Human Data Solution (HDS).” The extraction engine obtains unstructured data such as call logs, interview documents, customer feedback from different websites and social media platforms. HDS contains algorithms that classify this data into various categories, provides search engine and helps to gain insights about customer behavior. HDS combines unstructured data with structured data to provide meaningful insights. “If the clients choose cloud-based HDS solution, they can decide data sources for analysis using a simple user interface and get results of machine learning and predictive analytics in an easy-to-understand dashboard,” says Deshpande. Lending institutions, legal companies and market research firms are extensively using this solution.

To facilitate real-time data ingestion and analytics of streaming data, Ellicium provides a framework—Velocity Data Solution (VDS). The framework leverages Hbase NoSQL database, Kafka, and Spark for data storage, streaming and analytics. After periodically summarizing the data flowing from streaming sources, the predictive analytics algorithms are executed real time. Subsequently, the results derived from analytics are stored in a Hadoop database to enable business users to easily analyze them through visualization tools. “VDS processes terabytes of data and is time-series based—sequentially ordered data which is recorded on a regular time interval,” explains Deshpande.

The framework also has the capability to configure thousands of “on the fly” analytics functions on streaming data. “End users can configure analytics functions using a simple Graphical User Interface (GUI), without having to use any codes,” says Deshpande. VDS acts as a configurable data engine for big data to ensure good quality data is used for analytics. In addition, it provides variety of options for the users to visualize their streaming data.

Ellicium worked with a major telecom company that wanted to measure customer churn and produce marketing campaigns on customer behavior.

Ellicium’s solutions provide “customer insights” to our clients by analyzing massive unstructured and streaming data. Marriage of Big Data and Business Intelligence has helped us gain competitive advantage

However, the client found it difficult to handle the data volume from their streaming data sources which were running into multiple terabytes per day and was beyond the capacity of traditional RDBMS. Ellicium helped the client to leverage an architecture supported by Hadoop, HBase and Impala based VDS. This solution presented results in the form of easy-to-use dashboards and provided real time insights into customer behavior and enabled to company take actions based on predicted customer behavior.

Blurring the Lines between BI and Big Data Technologies

“The separation lines between big data and traditional business intelligence are blurring,” states Deshpande. Big data technologies such as Hadoop, Impala, Tez, and Big Query are increasingly being used to gain insights into large data sets by integrating with BI tools. Merging both these big data and BI technologies, Ellicium is developing a cloud based data warehouse for manufacturing and financial services industries. “This pre configured data warehouse can leverage large scale parallel processing capability of Hadoop,” says Deshpande.

Ellicium will be seen developing HDS as one stop shop for all text and unstructured data analytics needs in the coming days. Additionally, Ellicium will also be working on its on-demand cloud based Hadoop data warehouse with built in predictive analytics algorithms. The company will also introduce pay-as-you-go model for manufacturing and banking industries.


Pune, India

Kuldeep Deshpande, CEO

Focuses on solving complex data management issues pertaining to time-series and unstructured data