
Work
Projects
Developed an OpenCV pipeline to segment microscopy images and built hop-distance graphs to quantify spatial mixing across two adhesion systems. Measured 55% and 53% same-type neighbor fractions versus a 50% random baseline. Used Python, OpenCV, NumPy, SciPy, scikit-image, and Matplotlib with watershed segmentation.
Built a stochastic cell simulation in Python to model how bacterial cells grow, bind, break apart, and form multicellular clusters, using Gillespie-based event simulation to reproduce experimentally observed population behavior. Designed parameter-sweep experiments across repeated simulations, identifying binding conditions that reproduced experimental cluster sizes within ~5% and remained stable across a 10× change in starting cell population.

Built an African language learner covering 80+ phrases with pronunciation correction. Used Igbo foundation models from Hugging Face, automatic speech recognition, and text-to-speech.

Built a TypeScript/Next.js operations platform powering Connect Me’s tutoring programs for 5K+ users across 43 states and 12 countries, managing tutor pairings, scheduling, Zoom links, volunteer hours, and automated communications. Used TypeScript, Next.js, React, Supabase, Tailwind CSS, Zoom Web SDK, Twilio, Resend, and Upstash QStash.