Media Intelligence

Help Boost Your Business.

Monitor your brand health, online reputation, social media engagement & reach and your key topics with the same precision and structure of survey data, in real-time on PYXIS social media monitoring tool. 



Custom alerts & digests, Real-time alerts, Virality alerts, Daily digests, Custom digests, Mobile & Slack notifications, Custom PDF reports & Excel data export


PDF export, Excel export, Individual graphs export, Custom graph creation & Unlimited data storage


Big Data has become a differentiator in assisting companies to forecast and make effective strategic decisions to remain competitive.

Before data can be analysed, it needs to be collected. Based on the purpose  of  this data, the process of the collection of the data may differ. Typically there are three types of data that can be collected:


This is linear data, like the one that you see in the spreadsheet and are easier to evaluate. This kind data accounts to a small percentage of the data found today. 


It constitutes about 70 to 90% of today’s data and mostly comes in the form of text messages, videos, social media posts and more. Since it is unstructured, the analytical part of it makes it much more difficult and a lot of time is taken up in the processing.


This data has some tagging attributes and normally not easily read and understood by machine language. Examples are XML files or emails. 

Since this  data is to  be collected from a large number of sources e.g. cloud storage or from the operating system,  a dedicated server is required to save the uploaded files which are later extracted  for customisation scripting. 

After the data is collected, it has to be processed and sorted accordingly based on your requirements. Since this data is increasing daily or monthly, processing this data will become more challenging if we were to do it manually. This is where we require real-time processing.

Most of the data captured and stored are in BATCH form and this has been stored probably for a long time period of operations. These data blocks can be processed in batches but it will take a lot of man hours to see its output. 

Alternately, REAL-TIME ANALYTICS can solve this batch processing processor. The STREAMING PROCESSING of real-time analytics will resolve the issue of little-to-delay from the time the data is collected and process allowing businesses to make quick decisions. Streaming process is highly complex and more expensive.

Not all data would be relevant to the objective of the task which is undertaken. This is where cleansing or “scrubbing” has to be performed. All data that’s going to be used has to be standardised that is to format it the same. The migration of the data from the source file has to be performed and all duplicate and “no relevance data” has to be removed and purged. 

This process is equally TIME-Consuming and a lot of time is spent here for cleansing data for quality.

With data collected, processed and cleaned, it is then ready to be analysed. This is the final process where valuable information is extracted from the massive data storage. There are 4 type of  Big Data analytics: 

Diagnostic Analysis

This is to analyse why a particular problem occurred, where diagnostic visualization tools are used to drill-down to find out the root cause of the problem.

Descriptive Analysis

This is the most common form of big data analytics, where it gives an overview of the issue at a particular point of time. And this analytics are visualized through charts, graphs and reports.

Our Clients