With all the hype around big data, it is no wonder that often we forget to ask what exactly does this mean for our customers’ bread-and-butter business as providers of communications service providers, writes Shoma Chakravarty
There is no denying that new technologies have made it possible to affordably process vast amounts of diverse and changing data. At the core of obtaining value from big data, are the fundamental requirements that:
• The right questions are asked
• The answers are accurately surmised
• They can be translated into action
One might argue that generating value from every analytics use-case – be it finding genomic correlations with illness for pre-emptive healthcare, or detecting telecoms fraud, or increasing profitability, follows a similar three-step approach, says Shoma Chakravarty, is the chief technology officer at Telarix.
Measurable impact to shape your business performance
While it may be easy to philosophise about big data analysis, it is a tricky challenge for software to actually bring it to life, with clear and measurable impact. This is the challenge that Telarix has taken on as a provider of software automation for a dynamic customer base, with new entrants, an ever-emerging suite of services, competitive pricing pressures and a diverse regulatory landscape. As we made the leap to a big data platform enabling analytics-as-a-service, we focused on seven core principles to ensure the new technology provides our customers with quantifiable business value:
1. A system meant to provide value-generating insight, must first understand the end-to-end business process. This includes who the players are, what their role and responsibility is, how they interact with the data, and how their actions feed the lifecycle of the business.
2. The system has clearly defined the scope of use-cases for analytics, with quantifiable relevance to the business.
3. The system must allow for data exploration; allowing to ask pre-defined and ad hoc questions and exploring answers.
4. The system must provide an ability to control the data ingestion pipeline.
5. History is an old friend. The system must have the ability to utilise historical data for analytics.
6. The system must support action, based on the insight derived.
7. The system must be able to monitor, measure and report on its own performance.
The motivation for big data analysis must always be to facilitate meaningful insights while enabling fast decision-making with clear KPIs. Examples include: operational optimisations that increase margins or improve performance, insight about new opportunities for increasing revenues, or actions that rapidly arrest and mitigate revenue leakage due to abuse or fraudulent behaviour.
Beware the hype-appeal of a big data analytics platform that crunches data meaninglessly, burning CPU cycles and setting new records for IOPS (Input/Output Operations Per Second). In a disruption-prone market, where volumes are key to maximising profitability, and future services dictate an increasing demand on capacity and quality-of-service, we believe that having an analytics capability based upon the above principles will provide key and quantifiable differentiation to our customers.
The author of this blog is Shoma Chakravarty, is the chief technology officer at Telarix.
Comment on this article below or via Twitter: @ VanillaPlusMag OR @jcvplus
