Data is gold and every enterprise wants to monetize it. However, how to harness the power of data is still the biggest challenge. Some of the challenges are building infrastructure to store the data, identifying use cases, acquiring data sources, hiring skilled professional who can mine the data, and domain expertise to drive insights. Microsoft has recently released Azure Data Lake service on Azure Cloud Platform which addressed data storage and management problem by offering Big Data as Service. In this paper, we will discuss how Azure Data Lake can enable enterprise to accelerate Big Data adoption.
Ecommerce is a fast-moving and dynamic system. It is much different than regular commerce systems. The customers in ecommerce system have very small barrier to exit. It’s just one click away. On the other hand, once a customer enters a regular store various marketing methods (including visualization, touch, smell etc.) could be utilized to keep him engaged in the store.
It is very important in ecommerce to keep the customer engaged and interested. The statistics on ecommerce show that there is a big demand and large payoff for personalizing the offerings each of your customers gets when they visit your site.
The increasing complexity of supply chain in recent years warrants analytics for supply chain for any company that wants to
Digital and ecommerce is a very fast moving and dynamic system. Things could change in matter of hours. A tweet, a facebook update or a YouTube video can start a viral event in matter of hours. It’s very important for executive to stay on top of happenings.
Multiple events (promotions, coupons, marketing campaign, product launch, competitor action etc.) influence sales. It is important for an organization to accurately measure these influences. When these influences are modelled and tracked, the corporation can use them to increase the positive influences and reduce or eliminate the negative influences.
In this case study we demonstrate how Big data technology is used, to quickly analyze large volume of online activity (clickstream) data with large volume of off line sales (point of sales) data to gain business insights in an iterative manner.
Big data technology not only enable an organization to analyze large volume of data, but can also enable organization to analyze variety of data such as structured (sales, inventory), semi-structured (clickstream, log files) and unstructured (text, email, surveys) data. This capability creates unique competitive advantage for an organization in this digital age where information is abundant.
Apart from multi-channel attribution modelling, other types of analysis could also be performed on these types of dataset that includes; site optimization, product recommendation, campaign and promotion uplift models. These results are used as decisions support for marketing campaigns, and marketing mix model optimization.
Alerts can also be built on top of this analysis that can trigger email, text based notification to all the stakeholders when the result violates any pre-determined range.
Pharmaceutical manufacturing is a highly regulated industry. In this industry, traceability and genealogy are very critical to the manufacturing process. For best in class companies, automated, accurate, and timely traceability and genealogy calculations are critical.
One very large pharmaceutical client automated and implemented genealogy calculations for one of their products. These calculations are performed on a weekly basis, are complex in nature, and are a source of major performance and scalability concerns. This analysis uses distributed computing (using Hadoop and Pig UDF) to address the scalability and performance issues for genealogy calculations.