Software Release '26
byhand CLI
Makes language-model internals visible in terminal, HTML, and Excel, using supported GGUF models from Ollama, local files, or URLs.
I’m Jubair. I build tools that make AI models understandable.
Currently pursuing a PhD in Computer Science at the University of Colorado Boulder, I am advised by Prof. Tom Yeh. My work combines AI, visualization, and HCI to build tools that make AI models easier to see, understand, and use. Before my PhD, I earned an MSc in Computer Science at the University of Calgary, where I worked in the Visualization and Graphics Group.
Research Areas
Selected works
Technical Report '26
The Architect
Rebuilds deep learning math for forward pass and backprop inside MS Excel, so every step is visible in plain cells.
MLP Notebook
A small language for exploring the math, graph and code for a multilayer perceptron, inside the Jupyter notebook.
Scientific Data '24
ACCV Workshop '20
UIST Poster '22
AAAI Workshop '22UIST Adjunct '22
DIY Graphics Tab
A low-cost stand-in for a graphics tablet: a webcam tracks a pen moving on ordinary paper.
IEEE AICCSA '22
Book Cover Synthesis
Generates candidate book covers from summaries using text-to-image models and a curated dataset.
IEEE AICCSA '22
Shapes2Toon
Turns simple geometric sketches into cartoon characters with a GAN and an interactive web tool.
VMV '16
Icosahedral Maps
A method for converting complex climate model data into 2D structure to visualize at multiple levels of detail.
IEEE VIS Poster '15
IEEE ICTAI '20
toon2real
A method for converting cartoon images into realistic images using deep learning techniques.
VISAPP '22
Text-to-Expression
What you chat is what you get: a method for generating facial expressions from text descriptions.
Journey
2017 – 2022
Assistant Professor (Teaching-Focused, CSE)
Ahsanullah University of Science & TechnologyTalks & Workshops
-
Demystifying the Black Box
Artificial Intelligence in Clinical Practice Workshop, Kettering General Hospital
University Hospitals of Northamptonshire NHS Group, UKLed a hands-on session for clinicians, using matrix visualizations and numerical exercises to explain how neural networks process medical images, make predictions, and learn through backpropagation.
“With such clarity that even the most complex ideas felt intuitive.”
- 2025–2026
AI by Hand — Workshops & Paper Clubs
Virtual workshops co-hosted with Prof. Tom Yeh
Independently decomposed modern AI research into step-by-step matrix visualizations and numerical exercises, then led hands-on workshops using the materials I created.
20 events · 6 topics & formats · 1.7K+ registrations
Explore workshop topics
AI by Hand · Online
ND-Linear
Research by Ensemble AI
Worksheets for ND-LinearX-Attention
Research by Tsinghua, MIT, SJTU & NVIDIA · ICML 2025
Worksheets for X-AttentionSpiking Transformer
Research by UESTC & Liaoning Technical University · CVPR 2025
Worksheets for Spiking TransformerGeometry Self-Attention
Research by Nankai University (VCIP) · CVPR 2025
Worksheets for Geometry Self-AttentionBengali Edition
AI education in Bengali
The Architect — Demo
Created by me · Interactive neural-network blueprints
GitHub source for The Architect — Demo Watch demo -
Multiresolution Visualization of ICON Data
German Climate Computing Center (DKRZ), Hamburg, Germany
Invited by DKRZ to present my MSc research to climatologists and visualization researchers. I introduced multiresolution methods for exploring ICON climate data across scales using icosahedral maps and hexagonal box-spline wavelets.
Teaching
Five years as an Assistant Professor of Computer Science at Ahsanullah University of Science and Technology, where I designed and taught undergraduate courses—including Computer Graphics and Image Processing—for classes of 80–130 students.
2017–20224.83/5 teaching evaluation
CSE4203
Computer Graphics
Designed and taught a full lecture sequence spanning ray tracing, transformation matrices, viewing, rasterization, clipping, and Bézier curves.
Ray tracing teaching toolBresenham's algorithm teaching tool
CSE4204
Computer Graphics Lab
Created six hands-on WebGL and GLSL modules covering shaders, transformations, projection, texture mapping, animation, and lighting.
WebGL lecture playlistAlso taught Python programming at the University of Calgary and currently support Prof. Tom Yeh's Computer Vision courses at the University of Colorado Boulder.
Awards
- 2026 McLaren Futures Family Fellowship, University of Colorado Boulder
- 2026 Outstanding TA Award, University of Colorado Boulder
- 2023 Founder’s Fellowship, University of Colorado Boulder
- 2023 Service Award, University of Colorado Boulder
- 2022 Early Career Professional Development Fellowship, University of Colorado Boulder
- 2016 Travel Award, University of Calgary
- 2014, 2015 Research Award, University of Calgary
- 2014 International Recruitment Graduate Award, University of Calgary
- 2014 International Student Differential Fee Reimbursement Award, University of Calgary
- 2008 OIC Undergraduate Student Scholarship, Islamic University of Technology
Interested in working together?
I’m exploring research and engineering opportunities in AI, visualization, and HCI.