Tech, Research & Innovation
Exploring the intersection of technology, research, and entrepreneurship to create meaningful impact
Research
My research traces a path toward AI that's both thoughtful and useful – especially for people language models tend to overlook.
ML-Powered Robot Language Translator
Working with NIST's Cognition and Collaboration Systems Group to build a machine learning tool that translates programming code between industrial robot brands (ABB, KUKA, and others), so manufacturers don't have to manually rewrite code for every robot. I built the data pipeline – using RoboDK and Python to simulate robot tasks and automate dataset generation (cutting generation time from 15+ minutes to under 5 seconds per program, about 180x faster) – and helped fine-tune a transformer-based sequence-to-sequence model on the resulting multi-brand robot code dataset.
Tools & Technologies
Tags
Nearest Neighbor Nim: Combinatorial Game Theory Research
As an NSF CURM-funded undergraduate researcher, I worked on a research team analyzing Nearest Neighbor Nim (NNNim) – a variant of the classic Nim game where moves are constrained by an underlying graph's adjacency structure. We proved theorems characterizing winning positions on small graphs and built an NNNim calculator that uses recursive memoization and the MEX rule to compute game values for arbitrary adjacency structures, something no existing tool supported. I presented this work as a poster at the Maryland Collegiate Honors Council Conference.
Tools & Technologies
Tags
Calculus Behind the Magic: The Math of Facial Filters
My first dive into ML, written for my Calculus I Honors course. This paper breaks down how facial filter animations (like Snapchat's) work – using derivatives to track facial keypoint motion and linear approximation to model filter distortions – and connects these techniques to real applications beyond entertainment, including emotion analysis for mental health and accessibility. Written for peers with no coding background, it's where I first realized how the math I was learning connects directly to computer vision and ML.
Tools & Technologies
Tags
Future Research Interests
AI in Healthcare
Exploring ethical AI applications in medical diagnosis and treatment optimization.
Cybersecurity Automation
Developing automated threat detection and response systems for enterprise security.
Quantum Computing Applications
Investigating quantum algorithms for optimization and cryptography applications.
Sustainable Technology
Researching green computing and energy-efficient AI systems.
Work & Projects
Hands-on work spanning research labs, classrooms, and personal projects – each one an attempt to build something that actually helps people.
Mental Health Detection with NLP
An NLP project exploring how text-based signals can help identify indicators of mental health concerns – with careful attention to the ethical considerations of building tools in this space. In development for summer 2026.
Tools & Technologies
Tags
Wolof Clinical NLP
A personal community-driven project building NLP tools for Wolof, aimed at making healthcare information more accessible to Wolof speakers – a step toward my long-term goal of AI-powered tools that improve healthcare access in Senegal and beyond. In development for summer 2026.
Tools & Technologies
Tags
AiVise – AI-Powered Learning Platform
Hackathon project (1st place winner): built and deployed an AI-powered learning platform featuring an assignment checker, roadmap planner, and career tools, with full-stack integration across React, Next.js, PostgreSQL, Python, and REST APIs – prototyped and presented within 48 hours.
Tools & Technologies
Tags
Upcoming Projects
What's Next
Working toward contributing to open-source NLP tooling for low-resource languages as the Wolof clinical NLP project develops – more details coming summer 2026.
A Builder's Mindset
Before any of this – before CS, before NIST, before college – there was a small fashion business my sister and I started in Senegal in 2019. When we began wearing the hijab, we couldn't find clothes that worked for us: modest, but also stylish, in the styles we actually wanted to wear. So we designed our own – sketching pieces, sourcing fabric, working with a tailor to bring them to life. What started as solving our own problem became something friends wanted too, and for a few years, it was a real small business. We closed it in 2022 when we moved to the US, but the instinct it taught me – notice a real gap, design for it, build something tangible – has stuck with me ever since.
That instinct shows up in my Technology Entrepreneurship and Corporate Innovation minor, where every project has followed the same pattern: find a real problem, prototype fast, and figure out if it could actually work.
StyleSage
Personalized, sustainable fashion recommendations: a pitch for inclusive fashion tech, and in many ways the "scaled" version of what my sister and I were already doing by hand back in 2019.
NNNim
From research to playable product: turned our NNNim research into a strategy game designed to teach combinatorial logic without textbooks. We built an MVP (physical board, custom stones, and strategy guide) and piloted it with middle schoolers through Sonya Kovalevsky Day, with strong engagement from both students and educators.
IronLog
AI-powered fitness accountability: a Shark Tank-style pitch that pushed me through go-to-market strategy, unit economics, and competitive positioning for an app people would actually pay for.
I'm sitting on another idea right now that I'm not ready to share yet – but if the pattern holds, it won't be the last thing I build.
From Idea to Prototype
The fuller picture behind the NNNim story above – the MVP we built and piloted.
NNNim: Teaching Math Through Play
Turned our NNNim research into a playable strategy game designed to teach combinatorial logic without textbooks. We built a physical board prototype with a rulebook and strategy guide, then piloted it with middle schoolers through Sonya Kovalevsky Day – introducing them to game theory concepts through hands-on play. The pilot showed strong engagement from both students and educators, and we're exploring a digital version.
Tools & Technologies
Tags
Let's build something that matters.
Whether it's research, a project, or a conversation about using AI to make things better – I'd love to hear from you.