PhD Candidate @ Northeastern University

Shani C Spivak

My research lies at the intersection of data visualization, AI evaluation, and AI ethics. I develop methods to identify, surface, and characterize LLM errors in data visualization tasks. My recent work includes a taxonomy of LLM errors and an interview study of data workers on their use of LLMs.

Who I Am

Shani

I'm a PhD candidate in Computer Science at Northeastern University, advised by Dr. Melanie Tory (More about Melanie). My research sits at the intersection of AI, Ethics and Data Visualization.

Before starting my PhD, I got my B.S. in Electrical Engineering from Boston University and an MBA from NYU.

Other stuff I like: books, dogs, hiking

Even more about Melanie

My dog Maggie

NLP LLMs Visualization Policy and Governance AI Verification AI Literacy

Projects

Here are some of the projects I'm currently working on...

Screenshot of the AUXEL accessible data interface
Project 01

AUXEL

A natural language interface (NLI) to improve the accessibility of complex data tasks for blind or visually-impaired (BLV) STEM users. This research project supports the goal of increasing accessibility of spreadsheet data to BLV STEM data professionals and scientists.

Application User Study
Visualization of people's mental models of AI
Project 02

Visualizing People's Mental Models of AI

Investigating how how people conceptualize AI systems working. Through a series of workshops and surveys, participants visualize and build physicalizations of their “folk algorithms” or ”folk theories” about AI.

Workshops Survey
Code visualizations for surfacing LLM errors
Project 03

Surfacing LLM Errors: Can Visualizations Help Users Identify LLM Errors?

Using four different code visualization techniques in a user study, we explore whether these visualizations help participants identify errors in LLM-produced code.

User Study
IEEE VIS 2026 conference Boston skyline
Project 04

IEEE VIS 2026 OPC Assistant

Providing logistical, operational, and organizational support to the IEEE VIS Overall Paper Chairs.

Learn More

Selected Research

Human-centered Evaluation and Auditing of Language Models (HEAL@CHI) Workshop 2025
Fiction vs Friction: Challenges in Evaluating LLMs on Data Visualization Tasks
Shani C Spivak, Melanie Tory
Title and portion of paper
CHI 2025
The Many Tendrils of the Octopus Map
Eduardo Puerta*, Shani C Spivak*, Michael Correll
Teaser figure from the paper showing a big red octopus with tenticles in CHI locations

Let's Connect

Maggie

Maggie the dog
Maggie with a toy
Maggie sitting nicely
Maggie asleep with a smushed face
Maggie half on the couch
Maggie sleeping on a couch

This is Maggie, she takes up most of my couch and camera roll.