From theory to practice: Our journey as interns at SAS Institute – as analytics consultants

Before the Internship Started 

We are BI Norwegian Business School students of MSc in Business Analytics with a deep passion for data-driven analytics. What we love about data science is the journey towards achieving three significant kinds of results: discovery, insight, and innovation.

Data science has been an interdisciplinary field about processes and systems to extract knowledge or insights from data in various forms, either structured or unstructured, which is a continuation of some of the data analysis fields such as statistics, machine learning, data mining, and predictive analytics. Still, without the expertise of professionals who turn cutting-edge technology into actionable insights, Big Data is nothing. Now, more and more organizations are opening up their doors to big data and unlocking its potential—increasing the value of a data scientist who knows how to tease actionable insights out of gigabytes of data.

 The Connection between academia and industry

Being Business Analytics students at BI, we had an opportunity to get SAS Software Certification Young Professionals (SAS SCYP) and then also to complete an internship at SAS Institute - the leader in analytics – and to work closely with their client Wider?e, the largest regional airline operating in the Nordic countries. 

We were placed under the customer advisory department as Junior Analytic Consultants in Norway. The department is responsible for helping potential clients of SAS in determining which type of software package is best suited for them and also demonstrating the use of software in a use case.

During the Internship 

For implementing our conceptual knowledge in the new world of work, we used the skills and theoretical business knowledge taught in the Business Analytics study program. It enabled us to utilize academic concepts during our internship program. It is our know-how to apply the practicality of knowledge and associate the information with the situations you face in the real world of work. However, it is said that there is a vast gap between the linkage of classroom knowledge and practical work, but we would like to differ. Because whatever knowledge we imbibed in the classroom came to our aid during our internship, be it analytics, business studies or coding. We could optimize our participation in the internship program because we were highly motivated and confident of our theoretical knowledge that we gained in pedagogical discourse. 

Within the first few days, we were acquainted with colleagues, various airline terms in the industry, and the ongoing exciting projects. One of the things that we liked the most about our internship was the trust and freedom given to us as interns to choose how to proceed with the project that we were given.  We were grateful for our program as it prepared us for such situations. Why? Simply because we had the opportunity to experience the real data science workflow from scratch during our study program.

Allow us to briefly list down some of the workflow steps that we went through as this is what our foundation in Data Science is built on. 

1. Understanding the Business Problem

The project chosen was about customer segmentation. However, asking the right questions is very important for a Data Scientist. A lot of questions were raised in the beginning of the project to understand the real business problem. Essentially, our objective was to create segmentation to gain insights into customer behavior and attributes. Furthermore, it was important to us that this insight can be converted into actions and add value in the future.

2. Collecting Data Source

Excited with the new project, we started understanding the data better from a data dictionary and colleagues (basically walking around the office to ask questions on data sources). Understanding the data source is crucial as it lays the foundation to feature engineering. It is so important that it could affect the accuracy of the models that you are building in the later stage. The data was gathered from different sources. All these different sources were combined to create the customer table (also called an Analytical Base Table (ABT). 

3. Data Preprocessing

Data preprocessing is crucial as it transforms raw data into an understandable format so it can be fed to machine learning models. Research shows data preprocessing accounts for 80% of the work of data scientists. 

During our internship, we experienced why it is important to invest time in data preprocessing. 

Together with Wider?e, we agreed on the goal to create a table that can be used for numerous purposes. Consequently, we had to start at the data sources and create a pipeline using different data mining techniques. This task consumed a lot of time, as we wanted to make sure the table lived up to everyone's expectations. 

After a couple of weeks of working intensively with SAS Data Integration Studio, we reached this point and we could start creating different summary statistics to better understand Wider?e's customer base.

4. Building Models and Model selection

After some research, we decided to test different unsupervised learning techniques such as K-means, latent class, self-organizing maps, as well as RFM models. We learned that evaluating unsupervised learning techniques is difficult and close collaboration with Wider?e is crucial to come up with an appropriate customer segmentation that generates as much insight as possible for the company.

After the Internship - Reflections

The internship has reaffirmed our passion for Data Science, and we are grateful for the opportunity to continue this project together with Wider?e in the form of our master thesis.

The internship experience as Junior Analytic Consultants at SAS Institute was extremely valuable for multiple reasons. We could gain work experience while diving deeper into some of the concepts we learned during our Business Analytics Masters. Furthermore, the internship was extremely versatile, and we enjoyed the challenges as well as the responsibility given to us. The fact that two different companies and various stakeholders were involved made good and efficient communication crucial, especially as we took over the function as a middleman. Additionally, both of us are mainly working with open-source programming languages such as Python or R, which made it interesting to work with different tools such as SAS Data Integration Studio and SAS Viya.

To conclude, we would like to take this opportunity to thank both SAS Institute and Wider?e for the great experience we had, and we enjoyed working together with all of you.

 Further questions?

Thank you for reading. We hope that this article could give you some brief insight into a Data Science workflow as well as our journey at SAS Institute.

If you have any questions, feel free to add us on LinkedIn.

Abhijeet and Lorenz