Turning Data into Decisions: My Journey with SAS and AI in Banking

[Lucia_Biasi]

Building Responsible AI: How UNSW Students and SAS Are Shaping the Future of Data Science

In today’s data-driven world, building AI models is no longer just about achieving technical accuracy. It’s about balancing precision with ethical responsibility. The rapid adoption of AI has created new opportunities across industries, but it has also brought significant governance challenges. Ensuring that systems are transparent, fair, and secure is critical—especially when they influence decisions in lending, credit scoring, and fraud detection. Without proper oversight, biased algorithms can result in unfair treatment of customers, regulatory penalties, and a loss of public trust.

Devanga Kumarage - Customer Success Intern at @SAS | Final year bachelor of commerce and information...

For me, this reality came into sharp focus after earning my SAS Certification in Business Analytics. I was eager to apply what I had learned in a hands-on setting, and I found that opportunity through my capstone project at UNSW. Together with my team, I explored home loan datasets to predict defaults among borrowers—a challenge that required not only technical expertise but also a strong focus on fairness and interpretability.

Applying SAS Tools to Real-World Challenges

Using SAS Viya, I was able to combine insights from my coursework with the advanced capabilities of the platform. What struck me most was how accessible SAS makes AI and analytics: even those from non-technical backgrounds can quickly build and test models. Features like automation of data preparation, imputation, and transformation, as well as support for ensemble modeling, made the process efficient and powerful.

Our analysis revealed key predictors of loan default, such as the number of derogatory reports and delinquent credit lines. But beyond identifying risk, our goal was to make these insights understandable to all stakeholders—technical experts, business leaders, and customers alike.

To achieve this, we developed an interactive prototype in Figma. By entering customer information, users could see the predicted risk category, the likelihood of default, and the underlying factors driving the prediction. The prototype even generated tailored loan recommendations based on the assessment.

This layer of explainability was crucial. In banking, where lending decisions can significantly shape a person’s financial future, it isn’t enough for a model to be accurate. It must also be transparent and interpretable.

Beyond the Model: Building Confidence and Skills

This project was about more than technical application. Every two weeks, my team presented our progress to stakeholders, which pushed me to refine my communication and public speaking skills. Over time, I became more confident in explaining complex AI models to diverse audiences.

One of my proudest moments came during the ‘Girls in Tech’ event, where I introduced high school students to the possibilities of careers in data and technology. Sharing how SAS tools can make advanced analytics both accessible and impactful was a rewarding way to inspire the next generation.

My journey didn’t end there. I’ve continued to grow by participating in the SAS Viya for Learners Challenge, where I am exploring new and innovative applications of SAS Viya and SAS Viya Workbench.

SAS & UNSW: Partnering for Student Success

This work is part of a broader, ongoing collaboration between SAS and UNSW Employability. For the third consecutive year, we’ve partnered through the CDEV3000 Practice of Work and CDEV6000 Partnered Work Project.

Related: How students are using analytics to address poverty

Recently, we welcomed a talented cohort of UNSW students to our Sydney office, where they showcased their preliminary solutions against the stunning backdrop of our city views. During their presentations, they shared ideas, assumptions, and potential implementation challenges. They also received invaluable feedback from SAS experts including Jonathan Butow, Marife O. Heijster, and Gavin Long, which will help them refine their projects further.

These students will present their final solutions to the SAS Lane Cove office, demonstrating not only their enhanced technical skills but also the ability to deliver practical, industry-ready solutions.

Looking Ahead

This partnership continues to deliver value on multiple fronts. For students, it provides real-world experience, professional growth, and industry mentorship. For SAS, it sparks fresh ideas and innovative thinking, reinforcing our role in shaping the future of responsible AI and data science.

As we look toward the final presentations and beyond, we remain committed to empowering the next generation of data leaders—equipping them not just with technical skills, but also with the ethical mindset needed to build AI that truly serves society.

Want to get involved? Learn more about SAS and its academic programs.