Data Analytics consists of applying advanced analysis techniques on raw data to extract "hidden" information.
We chose to say the word "hidden" because this information and conclusions are usually dissimulated among a large amount of data. Mastering Data Analytics techniques help our customers increase their profitability and performance as they are taking better informed decisions.
Interesting fact about data analytics is that you can apply it to almost all and any enterprise department and service. In fact, every single one of them has some data that you can deeply analyse and extract insights from it to pursue continuous improvements.
Thanks to the Data Analytics Service, our customers are able to reduce their operational costs by improving their business processes. This service also provides interactive tools that can be used to take strategic decisions.
Data Analytics, different categories
Our service consists of multiple subcategories:
Descriptive Analysis: It is the observation of past events the company went through during a specific period of time. The goal here is to process this data, summarize it, present it in a easy format and have it consumed by business managers and owners.
Diagnosis: At this step, we aim to understand the "why part" of these events. In some cases we use external data and combine it with the company's internal data to further explain some events.
For example, if we try to explain a drop or a spike in sales for a tourism business, we can import weather and geopolitical data, combine it with company's internal data, try to find correlations between them and see if there are any causations effects.
Predictive Analysis: We can certainly not read into a crystal ball, but we can apply advanced algorithms and statistical techniques on historical data to understand some behaviours in a specific business field. Using these techniques can help us predict outcomes while considering the calculated confidence interval.
Prescriptive Analysis: The goal here is to build a model that can produce set of recommendations to act on in order to maximize company's performances.
The techniques in this case are similar to those used during the predictive analysis. However, using AI, recommendations are combined with company's business rules to automatically test multiple scenarios and strategic approaches. The results are then compared and presented to the end users while highlighting the best actions to perform.
Data Analytics, necessary steps
There are some common and basic steps yet very important to follow to successfully implement a Data Analytics project.
Grouping: We need to have a clear idea of how we want to group the data that we want to analyse (it can be based on age, gender, customer size, regions, etc.)
Data Collection: Data sources can be multiple and stored in different locations (local servers, web applications, third party services, etc.). It is important to determine as early as possible, the availability of these data sources and how easy we can have access to them.
Staging Areas: There are many ways and tools that can be used to extract data and store it (in a persistent or non persistent staging area) before we can start the analysis phase. We need to make sure of the solidity of this step because it heavily affects the performance and the resilience of the entire data analytics project.
Data Cleansing and Data Preparation: The Model's quality and performance depend directly on the quality of the ingested data. The common error we usually see out there in many data analytics projects, is when analysts undermine the necessity and the importance of the data cleansing and preparation phase. In fact, in most cases, this critical step is considered as annoying and time consuming. In addition, analysts chose this job because they like to analyse data and produce recommendations and findings, so, they unfortunately rush this step to jump to the analysis part and conclusions... The results can be damaging to the entire project.
Our Data Analytics team members have been tested and certified by Microsoft to provide a high quality level of service to our customers. We have developed deep experience in this field and work with proven methodologies & best practices to successfully deliver solid and resilient Data Analytics projects.
Our team work with different technologies such as: Azure for data collection, data storage, data cleansing and preparation. We also use Power BI to deliver the interactive front end tool to the end users.
Customer Cases - Microsoft Tools
Icahn School of Medicine at Mount Sinai
Genetics researchers in the Icahn School of Medicine at Mount Sinai are continuously looking for new ways and technologies, to develop their understanding about various chronic diseases such as Cancer and many others.
Researches in genetics fields have greatly increased during the last decades, bringing huge amounts of data that are now available to researchers around the world. The first challenge that Icahn School of Medicine team members have faced was to find a compute power resource capable of handling such amount of data. The research team consists of bioinformaticians and geneticists, they are seeking for proofs to find and better understand the links between our genes and cancer diseases. The goal was to find a technology that can allow them find these answers in the large amount of available data.
Using exome sequencing methods (more information can be found here), the team has quickly reached the maximum capacity of local machines / servers.
Mount Sinai took the decision to leverage the compute power and resources available in the Microsoft Cloud infrastructures. In addition, Microsoft offers a specific tool and capacity to cover their needs, it is called Microsoft Genomics and is meant to process very large DNA sequencing data jobs and tasks.
Thanks to Microsoft Genomics, the team is able to directly upload the large volume of data to the cloud, the tool can then uniformly realign everything and let the researchers start quickly to run their experiments. With Azure Data Lake Analytics, the team has access to unlimited storage disks allowing them to archive their findings without worrying about space and technology limitations.
XTO Energy is a subsidiary of ExxonMobil group and has multiple oil facilities in the Permian Basin (south-west USA). This region is known for being one the largest oil natural resources in the world.
Oil resources in this region expand over 235 thousands square kilometres. The climate is rough, hot, harsh, and unforgiving, making work difficult for technicians who have to drive many kilometres to go from an oil well to another one. Traditionally, field technicians had to visit each oil facility, take notes of its status and report all of the information to the central office at the end of the visits. XTO Energy decided to change this process.
These oil facilities were built many years ago. At that time, they were not built with sophisticated communication devices like the ones we can see nowadays. The company still had to find a way to efficiently collect data from the field to not only improve the overall performance but also the work conditions.
IoTs and Cloud capabilities. XTO Energy decided to take advantage of Microsoft IoT services using the Azure platform. With IoT devices, they could collect data from the field and manage it using Dynamics 365 Field Services.
Once the company started having enough historical data about its different facilities, it was time to start applying advanced analytics techniques to help take smarter and more informed decisions.
Using Azure Databricks, Azure Data Factory and Azure Data Lake, XTO Energy has now a strong and resilient data analytics solution.
We deliver tailored training sessions to organizations who are looking to develop internal skills for Data Analytics. These sessions include hands on labs using Azure Synapse and Power BI. Your employees will build skills that will allow them to create advanced data models to cover your organization's future needs.