Machine Learning Engineer • Generative AI • Multimodal Systems • Ethical AI
PhD Candidate in Computer Science
Passionate about building production-ready AI systems that solve real-world problems in healthcare, aviation, and scientific discovery. Experienced in fine-tuning LLMs (Llama-3, GPT-family), multimodal models (text + vision), and secure MLOps pipelines. Currently pursuing a PhD while publishing research and contributing to open-source AI initiatives.
- Generative & Multimodal AI projects (vision-language models, scientific data captioning)
- Healthcare AI (predictive diagnostics, EMR systems, brain disease detection)
- Aviation & Safety-Critical AI (fitness-to-fly certification, domain-adapted LLMs)
- Open-source contributions (Hugging Face, VergelAI, ethical AI tools)
- Research collaborations leading to publications
- Fine-tuned Llama-3-8B for aviation question-answering → +38% factuality (2025 preprint)
- Built production multimodal chatbot for scientific text + image processing
- Deployed healthcare models achieving 92–95% accuracy in diabetes & brain disease prediction
- Mentored junior engineers and taught ML to 200+ students
- Email: ekemini.thompson@outlook.com
- LinkedIn: www.linkedin.com/in/ekemini-thompson
- GitHub: github.com/EkeminiThompson (you're already here! 😃)
- Google Scholar: Ekemini Thompson
- ResearchGate: Ekemini Thompson
I love turning complex, high-stakes problems—like certifying pregnant passengers to fly or detecting Alzheimer’s early—into elegant, deployable AI solutions with sub-100ms latency!
🚀 Let's advance AI together—reach out for collaborations, research, or just to chat about the future of multimodal models!