What is inside the revamped IBM’s Watson ?
April 15, 2019
How Blockchain will embrace the future in Technology?
April 23, 2019

Future Trends and Insights on Data Analytics Platforms

This blog captures a Fresh take and an in-depth analysis and opinion from our esteemed Director Mr. Joseph Jayakumar. He is a renowned industry leader writing on trends of data analytics, IoT, IBM Watson, RPO, Reskilling etc. He has focused largely on the usability/agility aspect of analytics for effective adoption and reskilling across verticals in IT and at present is trying to highlight concepts like “Augmented Analytics”, “Augmented Data Management”, “Conversational Analytics”, “Continuous Intelligence”, “Explainable AI”, and “Persistent Memory Servers”.

While these concepts emphasize the smart yet continually evolving improvement of user experience of analytics/BI platforms from a multifocal technology perspective the adoption of Data analytics/Business Intelligence (BI) platform solely depends on three important factors namely,
- The Functionality/Features of Platform.
- Its Usability/Agility
- Manager’s Willingness to Integrate Analytics into Decision-Making flow from multiple perspectives supporting business decision-makers.

In BI platforms Data preparation tasks such as Data cleaning, profiling, enrichment, appending, transformation, master data management, metadata preparations etc are done manually. This procedure remains a bottleneck in the analysis process and we are trying to automate this process further to our existing clients. Though Augmented Data Management is used to target and address issues to automate this flow as much as possible either via Artificial Intelligence (AI) and Machine Learning (ML) techniques. Going ahead, this type of automated data management will be crucial in the data analytics flow in a wide spread adoption for enterprises. However in essence, “Augmented Analytics” refers to an intelligent and automated data analysis flow. The tool can configure itself to handle each and every component of analytics based upon the nature of data.

While most BI platforms have started migrating to this SaaS it will definitely need lot of iterations and re-designing in the coming future. This is essential as it has to adapt to user expectations and we are providing our client’s complimentary webinars and IT consulting sessions to get the best out of Augmented Analytics on IBM platforms to begin with. Most importantly aspect of data querying is as important as data analytics itself. This can sometimes acts as a hindrance if not done diligently for managers due to its syntactic complexity and can hinder finding useful insights/patterns from the existing data set.

Infact Gartner has identified that by 2020 nearly 50% of analytical queries will be generated by Search, Natural Language Processing (NLP) and Voice-Based Commands (as in chatbots). These will be called as “Conversational Analytics” which when integrated with BI/analytics platforms can enable managers to find better insights with greater ease.

Eg :- IBM Watson, Alexa

A live scenario can include a manager who needs to incorporate real-time data sources into their data warehouse for accurate insights, e.g. pulling customer reviews from websites and integrating with historical sales data to generate better promotional future predictions. However the aspect of “Continuous Intelligence” is a pattern which can integrate real-time analytics within a business operation. This enables live processing of current and historical data to generate more reliable predictions. Integration of contextual data from heterogeneous data sources outside the data warehouse of the BI platform will be an added advantage and essential for decision-makers to enable the adoption process seamlessly.

While predictions provided by BI platforms are generated by AI/ML based models they often lack interpretability necessary for convincing decision-makers. For example, functioning of a neural network-based model to predict a stock price across currencies globally is actually a black-box for a manager who plans to use it for effective portfolio management. This is one reason why managers find it difficult to deploy data analytics in their organizations. Addressing this issue can lead to “Better and Explainable AI” and can prove to be a game-changer towards day-to-day as well as strategic decision-making for decision makers. Lastly emergence of big data demands that BI platforms should handle large amounts of data for faster processing and better system agility. This is imperative for real-time analytic systems as it handles large heterogeneous contextual data and an entity called as “Persistent Memory Servers” should be installed to help in adoption of new memory architectures to cater to needs of business decision-makers simultaneously for the IT of tomorrow.


DISCLAIMER: The views expressed are solely of the author and Amstar Technologies shall not be responsible for any damage caused to any person/organization directly or indirectly by implementation of these platforms without prior consultation.

Comments are closed.