Recently, a US-based e-commerce business firm has been benefitted by SupraES while executing a new business analytics solution. This solution is not just cost-savvy but also influential to deal with the rising data analytical needs of the firm.
The US based e-commerce business firm has a dynamic and established user base across the country. One can well distinguish the vast business size of XYZ by the fact that it has over 250 thousand products on its official website, which serve the needs of more than one million users. As it had been mounting its business quite determinedly, it required an innovative business intelligence solution that can run as fast as its growing business. As we at SupraES, are already skilful in developing ascendable and customized business solutions, this e-commerce firm contacted SupraES.
Formerly, XYZ has employed a blend of Informatica and conventional MySQL databases for handling and upholding whole data. However, the business expanded, and now they wanted to enlarge the business analytics solutions too. On the contrary, mounting up the Informatica based solution while reviewing the present versions need huge investments that would not be a fruitful business affair. Putting this important alteration in the client’s business needs, we suggested it to build a new solution using Cloudera Hadoop, Spark and related technology pack. We told the firm that developing a fresh Cloudera Hadoop-based business solution would not just save the capitals but would also make the ultimate solution very much scalable and customizable flexible, eventually making the final analytics engine a Big Data Solution.
XYZ was formerly employing an Informatica based legacy business intelligence and data transformation solution. The terabytes of data produced by manifold streams of data generation channels were not well handled by the old solution. Moreover, this solution was relatively costly to gauge up. We had to develop a fresh solution that could tackle all these issues. So, the initial contest would be data migration without negotiating or pausing the data streaming, else real-time data would be misplaced.
To start with, we configured and setup the Cloudera Hadoop solution and also framed the latest data warehouse centered on the examined requirements of XYZ. Then, we integrated all sort of data streams into this solution while keeping a close watch on saving the entire real-time data covering clickstream, purchases, and lots more. Coming to the next step, we employed Scoop scripts to drift all legacy data from current Oracle data warehouse to the latest Hadoop Distributed File System based data warehouse clusters. Here, we built custom Scoop tool scripts so as to program single historical load transfer, comprising 15+ terabytes of tangled data into different HDFS clusters.
In any business solution, how to store big data efficiently is an important question that must be well responded. Broadly, it is of utmost importance to extract significant insights from the warehoused data. This US-based client was keeping various forms of data, such as user information, product inventory, product delivery, logistics, employees’ details and so on. Streaming, altering and then uploading such data to data warehouse well within time from such a vast data volume was a daunting task within itself. Besides, it was very important from business’ point of view to perform computerized ETL conversion on certain definite categories of data simply to fast-track the process of generating reports.
Hadoop data warehouses are super effective in keeping vast informational data. It also has another advantage, which is Apache Hive. Apache Hive is a data warehouse foundation developed over Hadoop that helps in summarizing the summarization, query along with big data analysis. This is what we used for XYZ. We take on Apache Hive over Hadoop warehouse to fast-track data enquiry. We also developed custom User Defined Functions (UDFs) so as to tackle trade-specific use cases and speed up particular data requests. Hive was supplemented by Apache Pig Scripts and execution scripts as per the client’s requirements, which assisted in automatic conversion and incoming data loading. All this facilitates several business reports.
There is no point of an efficient data analysis at the backend unless you have an effective solution for its representation. Same was the case with XYZ. The e-commerce firm was operating an obsolete version for Tableau and the team there was well skilled of using it. What they wanted was a freshly cooked solution that could be incorporated seamlessly with the Tableau solution. Also, they wanted us to update Tableau, design and develop new dashboards as well as reports around the tool.
Our first step to provide a solution for this challenge was to upgrade the Tableau solution to newest editions without harming the currently framed dashboards and structure of reports. Once we accomplished this, we incorporated latest Hadoop-based data analytics solution with Tableau through Tableau APIs. The assimilation is made at the virtualization layer, which allows real-time generation of reports and outlines of processed Hadoop data. Then, we designed and developed fresh dashboards and reports to display all new sorts of raw and processed data offered by the latest Hadoop solution. We rationalized entire set of reports and dashboards, thus enhancing their visual appearance and making them user friendly.
This high performance app allows users to calculate average treatment costs in a blink of an eye
The newly developed treatment price estimator enhanced leads via official website to more than 20%
Owing to the high quality insurance data collected via the app, the client has optimized its insurance offerings.
The app has become so famed that just this one has increased website's traffic by about 20%