Ayisha Tabbassum
This month’s spotlight features Ayisha Tabbassum, a dynamic researcher and one of the speakers at our WiADS 2024 event. Ayisha’s work bridges the cutting edge of AI, multi-cloud architectures, and data systems, focusing on vulnerabilities in machine learning models and optimizing data operations. Her insights range from surprising discoveries in cross-cloud efficiencies to the importance of holistic system security.
Ayisha shared her definition of Data Science as the engine driving digital transformation and highlighted exciting tools like LangChain, Hugging Face Transformers, and MLflow that she and her students are using to push the boundaries of AI applications. Looking ahead, she envisions a future shaped by privacy-preserving AI, explainable systems, and the evolution of unified data ecosystems.
Read Ayisha’s full responses to gain deeper insights into her impactful work and vision for the future of data science below!
1. What are your current research interests?
My current research interests are at the intersection of AI, multi-cloud architectures, and data systems. Specifically, I’m focused on exploring the vulnerabilities of machine learning models to adversarial attacks and developing strategies to mitigate these risks. Additionally, I am passionate about leveraging AI-driven insights to optimize data systems across multi-cloud environments, ensuring scalability, security, and efficiency. I also work on integrating Large Language Models (LLMs) into dynamic, real-world applications to solve complex challenges in education, healthcare, and enterprise systems.
2. How do you define Data Science?
I define Data Science as the interdisciplinary field that transforms raw data into actionable insights and decisions. By combining AI, cloud computing, and scalable data systems, it empowers us to uncover patterns, predict trends, and solve complex problems across industries. For me, Data Science serves as the backbone of digital transformation, leveraging technology to address real-world challenges and drive innovation.
3. Can you share an interesting or surprising result you’ve found in your data?
One of the most surprising findings emerged during a multi-cloud optimization project, where integrating AI algorithms to distribute workloads dynamically revealed inefficiencies in certain cloud service offerings. By analyzing cross-cloud latency and cost metrics, we identified opportunities to reduce operational costs by 25% while improving system reliability. Another fascinating discovery came from a study on adversarial attacks, which exposed vulnerabilities in data pipelines rather than just the machine learning models, emphasizing the importance of holistic system security.
4. Are there any interesting new tools or libraries you or your students have been using?
We’ve been actively using LangChain for building AI-powered applications with LLMs, MLflow for managing the end-to-end ML lifecycle, and Apache Iceberg for optimizing large-scale data lake operations. In the multi-cloud space, tools like Terraform and Anthos have been invaluable for automating deployments and managing hybrid environments. These tools have enhanced our ability to experiment, scale, and secure both AI and data systems. My students are particularly enthusiastic about exploring Hugging Face Transformers and Kubernetes-based ML platforms for hands-on AI and data engineering projects.
5. What are you most excited about in the field of data science in the next 5 years?
I’m incredibly excited about the convergence of AI, multi-cloud systems, and edge computing to revolutionize how data science solutions are developed and deployed. Over the next five years, I foresee advancements in privacy-preserving AI and federated learning, enabling secure and decentralized data processing across industries like healthcare and finance. Additionally, multi-cloud optimization will redefine scalability and cost-efficiency, while explainable AI will bridge the trust gap in decision-making systems. The evolution of data ecosystems with unified platforms will make data insights more accessible and actionable for businesses globally.
