Dr. Somali Chaterji (pronounced shoh-MAH-lee CHA-ter-jee) is
an Associate Professor in Agricultural and Biological Engineering (ABE) and the Elmore Family
School of Electrical and Computer Engineering (ECE) at Purdue University. She earned her Ph.D.
in Biomedical Engineering from Purdue, followed by postdoctoral research at UT Austin and Purdue
(Computer Science). With a dual background in bioengineering and computer science, her research
spans applied machine learning (ML) for computer systems (Best Paper, ACM SIGMETRICS
2022) and computational genomics (Best
Paper, ACM BCB 2015).
Her work applies ML to computer systems, earning her the NSF CAREER award (CISE-CPS, 2022), and computer systems to ML, leading to two patents and the founding of KeyByte, a cloud computing startup where she serves as CEO and co-founder. KeyByte specializes in optimizing ML workloads in cloud-hosted databases. She publishes in top systems (Usenix OSDI, ATC) and applied ML venues (CVPR, ECCV, KDD), with a research vision centered on enabling IoT devices to perform real-time analytics within latency, compute, and energy constraints.
Her research focuses on rightsizing ML—optimizing ML execution for accuracy, latency, energy efficiency, and interpretability—with applications in autonomous perception, scalable genomics, and AI-driven decision-making in digital agriculture and healthcare. In computational genomics, Chaterji develops ML-driven frameworks for capturing the continuum of cellular states, leading to more biologically meaningful clusters. Her work integrates self-supervised learning, variational embeddings, and perturbation analysis to improve interpretability and clinical relevance in single-cell analysis.
These innovations are integrated into real-world applications, including autonomous perception, scalable genomics, and AI-driven decision-making in digital agriculture and healthcare. Chaterji is a Co-PI for the NSF CISE Center CHORUS (2024), Co-PI of A2I2, the Army's Assured Autonomy Innovation Institute, and co-organized the NSF CPS PI meeting (2024). She also contributed to Purdue's WHIN project (Lilly Endowment, 2018-2024) on AI-driven innovation in manufacturing and agriculture and was selected for the NAE Japan-America Frontiers of Engineering symposium (2023).
Read more
Innovatory for Cells and Neural Machines
Read More
A podcast about domain-inspired machine learning and data engineering, focusing on genomics and IoT areas.
Spotify Apple Info
ICAN is an applied machine learning lab with two thrusts, Thrust 1 is IoT and Cloud Computing, and Thrust 2 is Computational Genomics.
Read More: ICAN's Thrusts, Learning, Engagement, Impact, and Vision.
Recap for the two thrusts of ICAN: computational genomics and edge analytics