About
About DecodedPapers
Welcome to DecodedPapers — where AI research meets engineering practice.
Most ML papers are written to prove a technique works at benchmark scale. I ask a different question: what can a practitioner actually extract and apply? Here you’ll find working implementations of architectures and techniques from recent research — built with standard tools, modest compute, and the constraints real engineers actually face.
Each post documents not just the code, but the decisions: what I kept from the paper, what I adapted, what broke, and why. The goal isn’t to reproduce benchmark numbers — it’s to give you something you can run, customize, and carry into your own work.
All implementations are available on GitHub. If a technique caught your attention in a paper but felt out of reach, this is where you find out whether it transfers.
About Me
Hi, I’m Mayer Antoine
I’m a data scientist, public health informaticist, and computer scientist fascinated by how technology solves real-world problems—especially applied AI/ML in public health and production ML systems.
-
🔭 I’m currently a Data Scientist at the CDC, helping public health professionals upskill and implement AI/ML solutions. My work spans predictive modeling, NLP, patient matching algorithms, big data analytics, data visualization and generative AI systems.
-
🔨 I build tools, proof of concepts, and tutorials that modernize health systems. Check out duplicategenerator (synthetic data generation for record linkage), cdc-text-corpora (public health text datasets), injury narrative coding (NLP for medical coding), and patient matching (probabilistic record linkage algorithms).
-
📚 I advance AI/ML capabilities across CDC through technical consultation, problem-solving, and creating learning materials that bridge research and real-world application.
-
🎓 Graduate of the Public Health Informatics Fellowship Program (PHIFP) and Georgia Tech’s MSc in Computer Science.
-
💡 I love projects at the intersection of AI/ML research and healthcare impact.
🛠️ Technologies I Work With
Machine Learning Transformers, PyTorch, TensorFlow, Scikit-Learn, Hugging Face, BERTopic, Langchain, OpenAI API and Agents SDK, Weights & Biases.
Cloud & MLOps Azure ML Services, Amazon SageMaker, Databricks, Docker, GitHub Actions.
Data & Development Python, JavaScript/TypeScript, SQL, Pandas, NumPy, FastAPI, React, Next.js, Jupyter.
📫 Connect
I’m best reached via LinkedIn. Always open to conversations about AI/ML in healthcare, public health informatics, or interesting technical challenges.
For more of my work, check out my GitHub.
Based in Atlanta, Georgia
🏆 Certifications
