Pankaj Agarwal

Pankaj Agarwal
Name	Pankaj Agarwal
Fields	Computer science; Algorithms (computer science)
Known for	Computational geometry; algorithm design; data structures

Pankaj Agarwal is an academic and researcher in computer science specializing in computational geometry, algorithms, and data structures. He has held faculty positions and research appointments at leading institutions and contributed foundational results that influence applications in computer graphics, geographic information system, robotics, and machine learning. His work bridges theoretical analysis and practical implementations used in both academia and industry.

Early life and education

Agarwal completed undergraduate studies at an institution with strong programs in mathematics and computer science, followed by graduate training culminating in a doctorate under advisers active in discrete mathematics and theoretical computer science. His doctoral research built on classical results from Edsger W. Dijkstra-influenced algorithmic analysis and techniques popularized by scholars affiliated with Carnegie Mellon University, Stanford University, and Massachusetts Institute of Technology. During his formative years he participated in seminars and workshops organized by ACM and SIAM panels, and presented early results at the Symposium on Computational Geometry and Foundations of Computer Science gatherings.

Academic and professional career

Agarwal has served on the faculty at multiple universities and research centers, collaborating with scholars from Princeton University, University of California, Berkeley, Harvard University, University of Illinois Urbana–Champaign, and international institutions such as ETH Zurich and University of Cambridge. He has been a visiting researcher at national laboratories and industrial research groups including Bell Labs, Microsoft Research, and IBM Research. Agarwal developed graduate curricula that interfaced with programs at National Science Foundation-funded centers and supervised doctoral students who later joined faculties at Columbia University, University of Toronto, and University of Washington. He has served on program committees for conferences like STOC, FOCS, SODA, and SoCG.

Research contributions and notable work

Agarwal's research advanced theoretical and applied aspects of computational geometry and related areas. He produced influential bounds for geometric data structures by extending methods associated with epsilon-net theory and VC dimension, integrating ideas from researchers at Yale University and University of California, San Diego. His contributions include output-sensitive algorithms for problems inspired by Delaunay triangulation and Voronoi diagram construction, and novel approaches to range searching building on techniques used in range tree and kd-tree literature. Agarwal introduced algorithms that improved worst-case and amortized complexity for dynamic geometric queries, connecting to paradigms developed at Princeton and ETH Zurich.

He co-authored papers that addressed nearest-neighbor search challenges using randomized incremental constructions and sketching methods related to work from Google Research and Facebook AI Research, enabling faster approximate queries in high-dimensional spaces used by image retrieval and point-cloud processing systems. His analysis of motion-planning primitives influenced research in robotics labs at Carnegie Mellon University and MIT CSAIL, and his techniques were adapted in computational pipelines at Esri and open-source projects stemming from OpenStreetMap-related tooling. Agarwal contributed to theory of geometric discrepancy and sampling with collaborators from University of British Columbia and Tel Aviv University, and co-developed algorithms for mesh generation that referenced results from Stanford and ETH Zurich groups.

Awards and honors

Agarwal's work has been recognized by nominations and prizes from organizations including ACM and SIAM, and by fellowships from national funding agencies akin to National Science Foundation and international scholarships comparable to Fulbright Program. He has been invited as a keynote and plenary speaker at flagship conferences such as SoCG, SODA, and international symposia hosted by Microsoft Research and Amazon Science. His students and coauthors earned awards at venues like ACM SIGGRAPH and the IEEE symposiums, reflecting the impact of his mentorship.

Selected publications

- Agarwal, P., et al., "Output-sensitive algorithms for Delaunay triangulation" presented at SoCG, addressing efficiency in planar and higher-dimensional triangulations. - Agarwal, P., et al., "Range searching and data structures for geometric queries" published in proceedings of SODA, building on kd-tree and range tree frameworks. - Agarwal, P., et al., "Nearest-neighbor and approximate search in high dimensions" in a journal linked to IEEE, with applications to computer vision and machine learning. - Agarwal, P., et al., "Mesh generation and geometric approximation" contributed to ACM Transactions on Graphics-adjacent venues, influencing finite element pipelines. - Agarwal, P., et al., "Discrepancy, sampling, and epsilon-net bounds" coauthored with researchers from University of Maryland and Tel Aviv University, advancing sampling theory in geometric contexts.

Personal life and legacy

Agarwal balances professional pursuits with involvement in academic service roles at institutions comparable to IIT-affiliated centers and international consortiums, fostering collaborations between North American and Asian research communities including networks tied to Indian Institutes of Technology and Tata Institute of Fundamental Research. His mentees populate faculties and research labs across Europe, North America, and Asia, continuing lines of inquiry in computational geometry and applied algorithmics. The body of work bearing his authorship endures in curricula, open-source implementations, and citations across venues such as ACM Digital Library, IEEE Xplore, and major conference proceedings.

Category:Computer scientists