Arvind Srinivasan

Arvind Srinivasan is a PhD candidate in Computer Science at Aarhus University, advised by Niklas Elmqvist at the Center for Anytime Anywhere Analytics. He studies attention in data analysis: what people look at shapes what they find, yet analytics tools keep no record of it. His attention-aware systems, on the web and in mixed reality, reflect that record back, so analysts can manage their own attention and others can follow, question and build on their work; the first two appeared at IEEE VIS and in IEEE TVCG. The interest began at the University of Maryland, where his work on how people notice useful analogies earned two Best Paper Honorable Mentions at ACM Creativity & Cognition. He now brings the same questions to the AI systems that do more of our looking: what did they attend to, what did they miss, and how would the people relying on them know?

Copy

I am a PhD candidate in Computer Science at Aarhus University, advised by Niklas Elmqvist at the Center for Anytime Anywhere Analytics. I study attention in data analysis: what people look at shapes what they find, yet analytics tools keep no record of it. My attention-aware systems, on the web and in mixed reality, reflect that record back, so analysts can manage their own attention and others can follow, question and build on their work; the first two appeared at IEEE VIS and in IEEE TVCG. The interest began at the University of Maryland, where my work on how people notice useful analogies earned two Best Paper Honorable Mentions at ACM Creativity & Cognition. I now bring the same questions to the AI systems that do more of our looking: what did they attend to, what did they miss, and how would the people relying on them know?

Copy

Arvind Srinivasan is a PhD candidate in Computer Science at Aarhus University, advised by Niklas Elmqvist, and one of the first hires at the Center for Anytime Anywhere Analytics (CA³). He studies attention in data analysis. What people look at shapes what they find, and it disappears the moment it happens. His systems reflect it back: seeing their own attention helps analysts manage it, and seeing someone else's lets people retrace how a conclusion was reached, question it and build on it. That trail lowers the bar for joining an analysis, a step toward analysis open to anyone, anytime, anywhere.

His thesis develops attention-aware analytics on the web, Meta Quest 3 and HoloLens 2, following attention from the analyst at work to collaborators sharing a mixed-reality space, an observer reviewing the session afterwards and a reader who receives only the findings. Built mostly on open web standards such as WebXR, the systems run in an ordinary browser or headset. The first two, Attention-Aware Visualization (2024) and HeedVision (2026), appeared at IEEE VIS and in IEEE TVCG, and related work has appeared at workshops of ACM CHI and ACM ETRA. As a visiting researcher at VISUS, University of Stuttgart, with Michael Sedlmair from October 2025 to February 2026, he studied where musicians look during band rehearsals.

The interest began with an MS in Human-Computer Interaction at the University of Maryland, where he worked with Joel Chan on how people notice useful analogies, including early studies of large language models as analogy generators; that work earned Best Paper Honorable Mentions at ACM Creativity & Cognition in 2023 and 2024. It now runs across ubiquitous analytics, collaborative immersive analytics, context-aware systems and human-centred AI. As AI systems take on more of the looking for us, he asks the same questions of them: what did they attend to, what did they miss, and how would the people in the loop know? Making attention visible on both sides is how he sees people keeping their cognitive autonomy, deciding for themselves where to look, what to trust and what to conclude. He reviews for ACM CHI, IEEE VIS and ACM Creativity & Cognition, and is a teaching assistant on the Data Visualization course at Aarhus University.

Copy

I am a PhD candidate in Computer Science at Aarhus University, advised by Niklas Elmqvist, and one of the first hires at the Center for Anytime Anywhere Analytics (CA³). I study attention in data analysis. What people look at shapes what they find, and it disappears the moment it happens. My systems reflect it back: seeing their own attention helps analysts manage it, and seeing someone else's lets people retrace how a conclusion was reached, question it and build on it. That trail lowers the bar for joining an analysis, a step toward analysis open to anyone, anytime, anywhere.

My thesis develops attention-aware analytics on the web, Meta Quest 3 and HoloLens 2, following attention from the analyst at work to collaborators sharing a mixed-reality space, an observer reviewing the session afterwards and a reader who receives only the findings. Built mostly on open web standards such as WebXR, the systems run in an ordinary browser or headset. The first two, Attention-Aware Visualization (2024) and HeedVision (2026), appeared at IEEE VIS and in IEEE TVCG, and related work has appeared at workshops of ACM CHI and ACM ETRA. As a visiting researcher at VISUS, University of Stuttgart, with Michael Sedlmair from October 2025 to February 2026, I studied where musicians look during band rehearsals.

The interest began with an MS in Human-Computer Interaction at the University of Maryland, where I worked with Joel Chan on how people notice useful analogies, including early studies of large language models as analogy generators; that work earned Best Paper Honorable Mentions at ACM Creativity & Cognition in 2023 and 2024. It now runs across ubiquitous analytics, collaborative immersive analytics, context-aware systems and human-centred AI. As AI systems take on more of the looking for us, I ask the same questions of them: what did they attend to, what did they miss, and how would the people in the loop know? Making attention visible on both sides is how I see people keeping their cognitive autonomy, deciding for themselves where to look, what to trust and what to conclude. I review for ACM CHI, IEEE VIS and ACM Creativity & Cognition, and I am a teaching assistant on the Data Visualization course at Aarhus University.

Copy

Arvind Srinivasan builds interfaces for working with data across screens, headsets and teams, and records what people attend to so that work can be traced and handed on. As a PhD fellow at the Center for Anytime Anywhere Analytics, Aarhus University, he builds these systems end to end for the web, Meta Quest 3 and HoloLens 2 and runs the user studies that test them; two are published in IEEE TVCG. He works mostly in TypeScript, React, WebXR, Python and Unity.

Copy

I build interfaces for working with data across screens, headsets and teams, and I record what people attend to so that work can be traced and handed on. As a PhD fellow at the Center for Anytime Anywhere Analytics, Aarhus University, I build these systems end to end for the web, Meta Quest 3 and HoloLens 2 and run the user studies that test them; two are published in IEEE TVCG. I work mostly in TypeScript, React, WebXR, Python and Unity.

Copy

Arvind Srinivasan builds interfaces for working with data across screens, headsets and teams, and records what people attend to so that work can be traced and handed on. As a PhD fellow at the Center for Anytime Anywhere Analytics (CA³), Aarhus University, where he was one of the first hires, he took attention-aware visualization from a single-user web prototype to co-located collaboration in WebXR on Meta Quest 3 and to asynchronous handoff between analysts. He wrote the gaze capture, per-mark visibility tracking and event-logged back ends himself, ran the user studies behind each step and published the results in IEEE TVCG. He also looks after the centre's website, branding and publication pipeline.

He is building heed, an open-source toolkit that records attention on any web page from cursor, viewport and webcam gaze, and has released caav, the toolkit from his HeedVision paper, which does the same for 3D objects. Before his PhD he earned an MS in Human-Computer Interaction at the University of Maryland, where he built a paper-processing pipeline and Elasticsearch-based analogical search as a graduate research assistant; that line of work received two Best Paper Honorable Mentions at ACM Creativity & Cognition. Earlier, as a UX intern at Tetrate, he redesigned the company's core platform for a new brand identity. He wants to build products that show what has been looked at and what has been missed, by a person, a team or an AI tool, so people can decide what to check before they act.

Copy

I build interfaces for working with data across screens, headsets and teams, and I record what people attend to so that work can be traced and handed on. As a PhD fellow at the Center for Anytime Anywhere Analytics (CA³), Aarhus University, where I was one of the first hires, I took attention-aware visualization from a single-user web prototype to co-located collaboration in WebXR on Meta Quest 3 and to asynchronous handoff between analysts. I wrote the gaze capture, per-mark visibility tracking and event-logged back ends myself, ran the user studies behind each step and published the results in IEEE TVCG. I also look after the centre's website, branding and publication pipeline.

I am building heed, an open-source toolkit that records attention on any web page from cursor, viewport and webcam gaze, and have released caav, the toolkit from my HeedVision paper, which does the same for 3D objects. Before my PhD I earned an MS in Human-Computer Interaction at the University of Maryland, where I built a paper-processing pipeline and Elasticsearch-based analogical search as a graduate research assistant; that line of work received two Best Paper Honorable Mentions at ACM Creativity & Cognition. Earlier, as a UX intern at Tetrate, I redesigned the company's core platform for a new brand identity. I want to build products that show what has been looked at and what has been missed, by a person, a team or an AI tool, so people can decide what to check before they act.

Copy