Doctor of Philosophy, Computer Science

University of Illinois Chicago, USA

• GPA (Current): 4.00/4.00
• Advisor: Fabio Miranda

Bachelor of Science, Computer Science & Engineering

Military Institute of Science and Tehcnology, Bangladesh

• GPA: 3.71/4.00 (Top 10th percentile)
• Thesis Advisor: Muhammad Nazrul Islam

Graduate Research Assistant

University of Illinois at Chicago, USA

• Advisor: Fabio Miranda
• Research area: Visualization and Visual analytics, Big Data Analysis, Applied Machine Learning

Graduate Teaching Assistant

University of Illinois Chicago, USA

• Courses:
• CS 422: User Interface Design and Programming
• CS 424: Visualization and Visual Analytics
• CS 425: Computer Graphics
  
  
• Duties:
• Served as TA for 4 semesters, conducting office hours for 40–60 students per semester.
• Graded assignments and exams, provided feedback on design and code.
• Assisted instructor in developing course materials and managing online learning platforms (e.g., Blackboard, Piazza).

PhD Data Visualization Intern

Epsilon, USA

• Created a pipeline to process ~1 million activity logs from the DiME business application, extracting user sessions to enable detailed behavioral analysis across user groups.
• Developed a pattern mining algorithm that computes recurring usage patterns, to understand user workflows and usage bottlenecks.
• Developed a visual analytics system that allowed stakeholders to easily uncover patterns in user sessions.
• Developed custom KPI based indicators (e.g., session duration, sessions per week) from the data.

Lecturer

Uttara University, Bangladesh

• Taught 7 core undergraduate computer science courses, consistently earning high student feedback ratings (4.5+/5) across multiple semesters.
• Supervised 6+ undergraduate students through their final-year projects and thesis submissions, leading to successful graduation.
• Created exams, assignments, lab materials, and contributed to curriculum development of courses.
  
  
• Courses Taught: Discrete Mathematics, Computer Peripheral Interfacing and Maintenance, Digital Logic Design, Computer Graphics, Object Oriented Programming, Design and Analysis of Algorithms, Data Structures.

Software Developer Intern

Solution Art Ltd, Bangladesh

• Completed a 2-month internship as part of BSc industrial training.
  
  
• Duties:
• Designed database schema for a hotel management system, improving data organization.
• Developed part of the frontend interface that supports intuitive booking for end-users.

Decision Support Tools for Sustainable Urban Planning and Public Health
(Poster)

• Developing oCUDS, a cross-domain visualization framework supporting urban decision-making across public health, transportation and climate domains.
• Designed with a multi-level dataflow model to enable flexible integration of preprocessing, visualization, and provenance tracking for diverse urban datasets.
• Integrating LLM-based natural language interface to allow end-users to query scenarios and decision-support tasks using intuitive, conversational prompts.

View Computation & Exploration in 3D Urban Environments
(Paper in review at IEEE TVCG)

• Developed a neural field based model to support view-based exploration of 3D urban environments, enabling tasks such as visibility analysis, solar exposure evaluation, and visual impact assessment of new developments.
• Achieved <10% prediction error in ~80% of regions across multiple synthetic scenarios.
• Outperformed KNN and random forest baselines in predicting semantic view composition, reaching 0.046 RMSE on a test set of 15,000 views and demonstrating robustness in low-data regimes.
• Engineered the model for real-time performance, achieving ~4 million views/second at scale with a 2.4 MB memory footprint, making it ~80× faster than traditional rasterization approaches.

Visual Gait and Motion Analysis
(Paper)

• Developed VIGMA, an open-access visual analytics system for gait and motion analysis, integrating computational notebooks and a Python library.
• Demonstrated support for multivariate gait data—kinetic, kinematic, and spatiotemporal parameters—using 120+ trials from healthy and stroke patients collected at baseline and 6-month follow-up.
• Validated the system with 5 domain experts from 3 research labs and 1 clinic, achieving 4–5/5 usefulness ratings and demonstrating improvements in error correction, disease tracking, and group comparison.

Bi-GRU Model for Automatic Gait Event Detection in Older Adults
(Paper)

• Developed a Bi-GRU-based automated gait event detection model achieving >97% accuracy and <14 ms mean error in both regular and perturbed walking.
• Analyzed gait data from 307 healthy older adults, demonstrating the model’s robustness across challenging perturbed walking scenarios where traditional force plate methods fail.

Navigating Large Dining Hall Spaces Considering Dietary Restrictions

• Designed and developed a React Native mobile app for UIC dining hall navigation, enabling personalized meal planning and efficient access to dietary-specific food options.
• Conducted user research with 6 participants and a comparative study against the existing “Dine on Campus” app, showing improved efficiency and user satisfaction.

Generative Model for Global Sunlight Access and Shadows
(Paper)

• Developed Deep Umbra, a novel computational framework utilizing a generative adversarial network to quantify sunlight access and shadows in urban environments at a global scale.
• 6x faster compared to the state-of-the-art accumulated shadow computation techniques.
• Open-sourced a comprehensive shadow dataset for over 100 cities, validated by a low RMSE (~0.06).

Trustworthiness of OpenStreetMap Sidewalk Data in US
(Paper)

• Conducted a comparative analysis of OSM sidewalk data across 54 major U.S. cities, revealing that 80% cities had less than 5% of sidewalk data available, highlighting severe data scarcity.
• Developed a trustworthiness index based on historical OSM edits to evaluate data reliability, revealing that even where sidewalk data exists, it is often unreliable—for example, in Chicago, only 24.4% of roads and 9.8% of sidewalk geometries had a trust index ≥ 0.5, with similar trends in Seattle and New York City.

COVID-19 Impact Analysis in Chicago Neighborhoods

• Built a random forest regression model to predict COVID-19 deaths per 1,000 residents across Chicago ZIP codes using sociodemographic features such as income, housing, ethnicity, and transit usage.
• Achieved a training error of 0.29 and test error of 0.78 deaths per 1,000, demonstrating strong predictive performance for neighborhood-level COVID-19 impact.
• Applied principal component analysis (PCA) to explore variation in death rates across ZIP codes, revealing no clear patterns due to mixed urban, suburban, and rural contexts.

Chicago Taxi Ridership Visualization Tool

• Developed an interactive visualization tool to analyze 2019 Chicago taxi ridership trends, using 7GB of trip data hosted on UIC's EVL shiny-server.
• Optimized for large-screen displays and fast load times by splitting data into subfiles and pre-processing with Python scripts, enabling real-time exploration of usage trends. Discovered key insights such as low weekend ridership, rush hour spikes in the Loop, and a high concentration of short-distance, short-duration trips within central Chicago.

Mobility-Flow Query Approximation using NeuralCubes

• Developed an in-memory model using NeuralCubes to approximate spatiotemporal mobility-flow queries with high accuracy.
• Achieved less than 2% absolute error in query approximation while maintaining a minimal memory footprint of 114 KB.

Autism Spectrum Disorder Prediction Model and Mobile Application
(Paper 1, Paper 2)

• Undergraduate thesis.
• Built a novel random forest ML algorithm achieving 92%+ accuracy on the AQ-10 dataset for autism prediction and evaluated performance on both AQ-10 and real-world datasets.
• Developed a mobile application to deploy the model for accessible, real-time screening.

IoT-based Assistive Tool for Alzheimer’s Patients
(Paper)

• Undergraduate research project.
• Designed and prototyped an assistive system with a mobile app to support Alzheimer’s patients and caregivers through health monitoring, medication reminders, item tracking, and location monitoring.
• Conducted a focus group study with 15 participants (students and faculty); 87% found the system functionally accurate, and 100% rated it easy to use, validating usability and system effectiveness. Integrated multiple assistive features into a single, unified platform—a novel contribution over prior fragmented solutions—with potential for future enhancement as wearable and offline-capable modules.

Developing an Open Computational Framework for Decision Support Across Transportation, Weather, and Public Health

• Poster presented at Sustainability Research + Innovavtion Congress 2025.

Visual Analytics Approaches for Facilitating Explainability of Graph Neural Networks

• Ph.D. Qualifier Exam, 2023.

Crowdsourcing and Sidewalk Data: A Preliminary Study on the Trustworthiness of OpenStreetMap Data in the US

• Paper presented at ASSETS’22 Workshop on The Future of Urban Accessibility, 2022.

An Intelligent Assistive Tool for Alzheimer’s Patient

• Paper presented at IEEE ICASERT conference, 2019.

A Machine Learning Approach to Predict Autism Spectrum Disorder

• Paper presented at IEEE ECCE conference, 2019.

• Merit Scholarship for academic performance, Military Institute of Science and Technology

• Dean’s List Award (two consecuctive years), Military Institute of Science and Technology

• Paper reviewer for PacificVis 2024, EuroVis 2023-2024, IEEE VIS 2022-2024.

• Volunteer for CAVE3 demos hosted by EVL, UIC.

• Vice President of Media of Bangladeshi Student Association at UIC.

• IEEE VIS Satelite Event (2021) volunteer. Held at EVL, UIC.

• Class Representative, Department of CSE at Military Institute of Science and Technology.