avatar

Suhas Thejaswi

Researcher
Aalto University
firstname.lastname@aalto.fi


About Me

I am a Postdoctoral Researcher at Aalto University, Finland. I received my PhD in Computer Science from Aalto University, where I was fortunate to be advised by Prof. Aristides Gionis. I previously worked as a Postdoctoral Researcher at the Max Planck Institute for Software Systems, Germany. Earlier, during my master’s studies, I had the opportunity to work as a research assistant at Aalto University. Before starting my master’s, I worked as a (research) engineer at Nokia, Motorola, and Ericsson.

Research Interests

My primary research interest lies in the design of algorithms with provable theoretical guarantees, with an emphasis on responsible computing in automated decision-making (or decision-support) systems. To me, responsible computing is not only about mitigating systemic biases, unfairness, and discrimination, it is a broader commitment that requires us to look beyond empirical performance and ask foundational questions: What are we computing and why? Is the objective that we compute meaningful in its context? It further demands foresight: to anticipate the downstream consequences of replacing human judgment or existing institutional processes with algorithmic decision-support systems. My work is guided by these perspectives, striving to build theoretically grounded, socially aware, and future-ready algorithmic solutions that not only solve today’s problems, but are also resilient to the societal and technical challenges of tomorrow.

My past research has focused on designing scalable algorithms for societal challenges, including modeling epidemic propagation and developing algorithms for fair variants of unsupervised machine learning problems such as clustering, dimentionality reduction and feature selection, as well as addressing social choice questions such as matching, multi-winner voting and committee selection. More recently, I have been working on the design of algorithmic decision-support systems that promote effective human-AI interaction—so that humans and algorithms together can perform better than either could alone—with the goal of achieving human-AI complementarity. In this context, I am also investigating how biases perpetuate when humans use algorithmic predictions to make final decisions, and how such biases can be mitigated.

Recent News