About Me
I am Tresor Yubahwe, a Computer Science student at Princeton University and a Software Engineering Resident at HeadStarter, driven by a passion for AI, data science, and creating impactful solutions for underserved communities.
Skills
Java, Python, React, Firebase, PyTorch, and TensorFlow.
Projects
Personalized Learning System
Developed an AI-powered system to improve student engagement rates by 30% using Hugging Face and PyTorch. Applied natural language processing (NLP) for content embeddings and trained models with CUDA-optimized GPUs, enhancing learning outcomes in Rwandan schools. Future plans include integrating adaptive feedback loops to further individualize learning paths.
USSD Chatbot
Built a USSD-based chatbot integrated with Twilio's WhatsApp API for delivering sexual and reproductive health education to underserved communities. Leveraged Flask for backend operations, reaching 65+ women with no smartphone access. Future enhancements include adding multilingual support and expanding the chatbot's functionality to include mental health resources.
Pentagram Project
Collaborated as a Software Engineering Resident to optimize user workflows for recruitment at HeadStarter. Designed and implemented advanced analytics with modern UI principles to streamline processes, improving user efficiency. Future goals include integrating AI-driven recommendations for dynamic insights into hiring trends.
Pantry Management System
Designed and implemented a pantry management system with Next.js, Material UI, and Firebase for real-time inventory updates. Integrated GCP Vertex AI and AutoML for intelligent image classification to categorize pantry items. Future plans include expanding to support multi-location tracking and detailed nutritional analysis.
AI Answer Engine
Developed an AI Answer Engine leveraging Hugging Face Transformers for real-time domain-specific question answering. Enhanced performance by fine-tuning transformers for precision and speed. Future upgrades will incorporate multilingual support and broader datasets to improve contextual understanding across industries.
Operational Strategy at Zerox Tech
As Chief Operating Officer at Zerox Tech, implemented operational strategies to scale company processes using data-driven solutions. Conducted efficiency analyses and introduced AI tools for streamlined operations. Future plans involve exploring predictive analytics for optimizing long-term business planning.
Princeton AI Alignment Lab
Conducted research on ethical implications of AI, focusing on reinforcement learning and AI safety techniques such as ELK (Eliciting Latent Knowledge) and mechanistic interpretability. Collaborated with peers to analyze readings on topics like reward misspecification and scalable oversight. Applied findings to propose scalable frameworks for safe AI deployment. Future goals include publishing findings and extending this work to practical AI ethics courses.
Here for You (H4U)
Founded and led Here for You, an initiative addressing learning poverty in Rwanda through digital literacy and civic entrepreneurship bootcamps. Delivered 1,800 hours of digital literacy training and facilitated civic projects for underserved students. Achieved national recognition for implementing programs resulting in a 300% increase in ICT engagement. Future plans include developing an online platform to scale H4U’s impact globally.