FORMCEPT: Expediting the Process of Data Driven Decision Making
Conventional data warehousing technology focuses on unifying the internal data, which is predominantly transactional, and derive business intelligence out of it. This is not effective in the current business scenarios because transactional data only accounts to 14 percent of the available data. The remaining 86 percent is unstructured data that is valuable as well. Unless both structured and unstructured data, whether it is internal or external, are unified, it is challenging to build a holistic view of the business problem. Further, Unified data view and consistent insights need orchestration of variety of technologies that is of prime concern for the Enterprises. FORMCEPT, headquartered at Bangalore, brings a solution to this problem in the form of a state-of-the-art “Unified Data Analysis platform” called MECBOT. MECBOT is an effort to make content analysis accessible to everyone, be it an enterprise, an individual or a device. “The MECBOT platform dramatically reduces the pre-processing time involved in the analytics cycle and makes the data available in consumable form for further analysis,” says Suresh Srinivasan, Co-founder, FORMCEPT. MECBOT not only provides a unified interface for storing and querying datasets but also provides batch processing, stream processing and interactive analysis out of the box. It is built on top of NoSQL technologies, like- HDFS, HBase for storage, systems like Map-Reduce, Spark, Shark and Storm for data processing and utilizes the power of graph database, key-value store and index store to provide a seamless experience for the end users. The platform incorporates the concepts of lambda architecture and can be deployed either On Premise or on Private Cloud.
Redefining data analysis
FORMCEPT believes that unless the data is unified, standardized and semantically harmonized with minimal noise, it is impossible for downstream analysis to derive consistent insights. With that belief, MECBOT is designed to stand out amongst other data analysis platforms owing to its capability to maintain a managed, semantically enriched data lake which enables each and every individual to access the data and perform various analysis on their datasets.
A major distinguishing features of MECBOT are
1) FactorDB, innovative metadata handling methodology, which sits on top of proven NoSQL technologies, like key-value stores.
2) Enrichment of content using a Knowledge base built on the concepts of Linked Data to identify precise context and extend across domains.
3) Seamless connection of data across multiple sources, using the patent pending Data Folding™ techniques to discover the relations and patterns. 4) Text Analysis, Graph based analysis and Statistical analysis can be done at a scale without worrying about Volume, Velocity and Variety
5) Scales horizontally and vertically on commodity machines.
It is due to this innovative approach followed by FORMCEPT that it has been recently awarded with 2014 New Product Innovation Award in Data Analysis Platforms by Frost and Sullivan. Currently FORMCEPT is working with telcos, retailers, global recruitment companies and sporting media giants across the globe on various pilots to create unified data assets, analyze and gain meaningful insights.
Panel
Unified Data Analysis
MECBOT addresses the creation of unified data assets efficiently. Its Data Ingestion Module provides Visual Interface where one can configure various Structured, Unstructured and Semi Structured data sources. Once the data is configured, various transformations on the data can be defined visually. The data is stored efficiently in FactorDB, Innovative Meta data handling methodology, which sits on top of proven NoSQL technologies, like key-value stores. While the data is getting stored, FORMCEPT applies advanced mathematical concepts and advanced NLP techniques to create DataFolds™ across the unified data assets. Datasets stored in FactorDB are available for various analysis either through APIs or through SQL interface or “Free textual search.” Data Scientists/Analyst can also do exploratory analysis to identify some patterns and can build models using well know languages like- R. Once the model is tested, it can be imported as PMML into MECBOT to run it at scale. The entire platform can be deployed in private cloud of choice or on premise and it scales elastically both vertically and horizontally.
3) Seamless connection of data across multiple sources, using the patent pending Data Folding™ techniques to discover the relations and patterns. 4) Text Analysis, Graph based analysis and Statistical analysis can be done at a scale without worrying about Volume, Velocity and Variety
5) Scales horizontally and vertically on commodity machines.
MECBOT is capable of maintaining a managed, semantically enriched data lake which enables each and every individual to access the data and perform various analysis on their datasets
It is due to this innovative approach followed by FORMCEPT that it has been recently awarded with 2014 New Product Innovation Award in Data Analysis Platforms by Frost and Sullivan. Currently FORMCEPT is working with telcos, retailers, global recruitment companies and sporting media giants across the globe on various pilots to create unified data assets, analyze and gain meaningful insights.
Panel
Unified Data Analysis
MECBOT addresses the creation of unified data assets efficiently. Its Data Ingestion Module provides Visual Interface where one can configure various Structured, Unstructured and Semi Structured data sources. Once the data is configured, various transformations on the data can be defined visually. The data is stored efficiently in FactorDB, Innovative Meta data handling methodology, which sits on top of proven NoSQL technologies, like key-value stores. While the data is getting stored, FORMCEPT applies advanced mathematical concepts and advanced NLP techniques to create DataFolds™ across the unified data assets. Datasets stored in FactorDB are available for various analysis either through APIs or through SQL interface or “Free textual search.” Data Scientists/Analyst can also do exploratory analysis to identify some patterns and can build models using well know languages like- R. Once the model is tested, it can be imported as PMML into MECBOT to run it at scale. The entire platform can be deployed in private cloud of choice or on premise and it scales elastically both vertically and horizontally.
