Adam Coscia

I am primarily motivated to help people make smarter decisions with their data. I tackle this problem by developing interactive visual interfaces that externalize and augment analytical reasoning between humans and machines.

Classically, machines automate computational tasks while humans provide judgement and guidance. Today's advances in AI and robotics are shifting that balance. Consider teams of analysts offloading data mining to swarms of agents, students incorporating personalized AI feedback on an essay before a single teacher reads it, operators chatting with a robot teammate that refused an order. These upheavals require us to examine age-old questions from a fresh lens. How do we enable bi-directional human-machine communication? How do we represent and interrogate a distributed reasoning process? How do we compare competing lines of inquiry across individuals and teams?

To answer these questions, my research group combines techniques from visual analytics, human-centered AI, and data science to build interactive interfaces that help people and AI make sense of complex data together. We primarily engage in iterative co-design and longitudinal field deployments of our systems with real domain experts, demonstrating the application of our methods to pressing challenges across diverse fields including AI in education, intelligence analysis, and human-robot teaming. Examples include visualizing how an AI grades a student's essay, tracking whether agents maintain user goals over multiple turns, and tracing what caused an autonomous robot teammate to not follow orders.

Our work has contributed new open-source software, empirical studies, and design guidelines to the visualization and HCI communities across three areas of research:
- Working with AI: Can showing analysts visual traces of their own analytic behavior make them aware of their biases? How do analysts validate evidence generated by AI about a collection of documents? (Lumos, Left Right and Gender, VisPile, OnGoal, Integration Guidelines)
- Learning from AI: What do LLMs actually learn during training, and how do learned associations surface in downstream tasks? How do engineers building AI-powered Edtech understand and trust the scores AI assigns to student writing? (KnowledgeVIS, iScore, Tutor Provenance, LLM Prompt Design)
- Creating with AI: How do we make AI's contributions to a co-written draft visible to human authors? How do we make compositional reasoning in GenAI steerable? (DraftMarks, Visual Concept Blending)

Adam Coscia is a Tenure-Track Assistant Professor in the School of Computing at Stevens Institute of Technology and a member of the Stevens Institute for Artificial Intelligence (SIAI). He received his Ph.D. in Human-Centered Computing from the Georgia Institute of Technology under Dr. Alex Endert, and holds a B.S. in Physics from Stevens. His research combines techniques from visual analytics, human-centered AI, and data science to build, deploy, and study interactive systems for reasoning and decision-making over complex data. He has designed and deployed his tools for intelligence analysis, education, healthcare, and robotics, in collaboration with both industry and government researchers, analysts, and scientists at the National Security Agency, MIT Lincoln Laboratory, NASA Jet Propulsion Laboratory, Adobe Research, the NC State Laboratory for Analytic Sciences, and the Monterey Bay Aquarium Research Institute. His work appears at premier HCI and visualization venues including ACM CHI, ACM UIST, ACM IUI, IEEE VIS, and IEEE TVCG, and has been recognized with a Foley Scholar Award nomination, Georgia Tech's President's Fellowship, and three best poster awards. He serves as Technology Chair and on the Program Committee for IEEE VIS 2026.

Research Assistant
Graduate Research Assistant, Visual Analytics Lab, School of Interactive Computing, Georgia Institute of Technology, Atlanta, GA, 2020-2026
Undergraduate Research Assistant, School of Information Systems and Architecture, Stevens Institute of Technology, Hoboken, NJ, 2018-2020

Internships
Artificial Intelligence Group, MIT Lincoln Laboratory, Lexington, MA, 2025
AI Experiences Lab, Adobe Research, San Jose, CA, 2024
Human-Interfaces Group, NASA Jet Propulsion Laboratory, Pasadena, CA, 2023
Data to Discovery Program, NASA Jet Propulsion Laboratory, Pasadena, CA, 2021
Center for Data Science and Artificial Intelligence, New York Life Insurance Company, New York, NY, 2020
Center for Data Science and Analytics, New York Life Insurance Company, New York, NY, 2019
Department of Physics and Astronomy, Katholieke Universiteit Leuven, Leuven, Belgium, 2017

Somethings, Technical Advisor, 2026-present

Assistant Professor, School of Computing, Stevens Institute of Technology, 2026-present

Research
Best Poster Award, CRIDC, 2025
Foley Scholar Award Finalist, GVU:IPAT, Georgia Tech, 2024
Best Poster Award, CRIDC, 2023
Best Poster Award, CRIDC, 2021

Teaching
Distinguished Teaching Assistant Award, Stevens Institute of Technology, 2018

Honor Societies, Fellowships, and Others
Sigma Xi Scientific Research Honor Society, Sigma Xi, lifetime member, inducted 2025
Sigma Pi Sigma Physics Honor Society, American Institute of Physics, lifetime member, inducted 2019
President's Fellowship, Georgia Tech, 2020
Alfred M. Mayer Prize, Stevens Institute of Technology, 2020
Presidential Scholarship, Stevens Institute of Technology, 2016

CS 538 Data Visualization, Spring 2027