[vc_row speed=”0.20″ row_seperator=”no-seperator” top=”0″ bottom=”0″][vc_column width=”1/1″][vc_column_text]Note: This guest post was written by Mike Prorock, Director, Emerging Technologies with Bardess’s partner Attivio.
One of the difficulties in any data-driven solution is unifying disparate sources of data and content. The reality stands that the vast majority of data that exists in the wild is semi-structured at best. In working with a customer of ours focused on Call Center analytics, Bardess was posed with an interesting dilemma: how do we unify application-generated, customer, and multimedia content for consistent, high-value analytics?
Capture the Voice of the Customer
One of our customers, TopBox, had created a novel and industry-leading solution to identify and capture data within customer interactions around operational and business problems as they arise, and to gain actionable insights into those problems. By using TopBox’s solution, customers can target a subset of interactions, capture hidden data using a patent pending technique and analyze unstructured data in single end-to-end solution. The outcome for the customer is the ability to clearly identify root causes that were simply undiscoverable before, and enable solutions to problems to be deployed across their contact centers in a fraction of the time previously required.
Spotfire proved a perfect tool for visualization, workflow control, and analysis within the solution, but it still needed to be able to access the data across the variety of sources. This is where Attivio was introduced to solve the data source discovery gap.
Ingest Disparate Data Sources
Bardess was able to rapidly deploy custom connectors tailored around TopBox’s needs to pull their disparate data together for analysis in Spotfire. Upon implementation, Attivio began ingesting content from:
- MongoDB, the application database
- Unstructured content such as customer textual information from S3 and file system locations
- Relational tables and tabular flat files
MongoDB posed an interesting dilemna. By nature, MongoDB is document-oriented like Attivio, but frequently stores data in a more structured way. To gain optimal value from the data, and to respond to changing schemas and document structures within Mongo, the connector needs to be tailored to auto-discover changes and to preserve nested relationships between data so that the data stored in MongoDB could be accurately visualized in Spotfire and recreated in a tabular format as needed upon retrieval from Attivio.
Unify Data Sources to Feed the BI Tool
Query time joins in Attivio let us unify this data across the sources of data for Spotfire to consume via access controlled information links in an optimal format for visualization and analysis. Further, users of the system were able to rapidly search text and other unstructured data within the Spotfire analyses to limit the visualized data only to areas of interest. Keyword and sentiment extraction brought additional value to the unstructured data providing highly important data points for analysis.
Without the use of Attivio, a highly complex system would have had to have been built to address the problem, and the time to market would have been greatly increased, along with the overall cost of the solution. Two developers were utilized over the space of a few weeks to dynamically extract, ingest, and add value to the content in Attivio for the Spotfire analysis. Attivio became the cornerstone to unify, search and retrieve relevant data, enabling TopBox users to identify problems, get to the root causes of those problems, and deploy solutions more rapidly than they ever have before.[/vc_column_text][/vc_column][/vc_row]