Daniel Cavadia

AI/ML Engineer

Computer Scientist with minor in AI, specializing in medical AI and machine learning applications. I develop advanced CNN models for healthcare diagnostics, build intelligent systems with LLMs and RAG architectures, and collaborate with medical and financial institutions on cutting-edge algorithms.

PINNED PROJECTS

The pinned projects displays my work for easy reference.

MelanoScope AI

Bachelor's thesis research developing advanced convolutional neural networks for automated melanoma detection. Achieved high accuracy in medical image analysis using NFNet deep learning architecture for early skin cancer identification and classification.

TensorFlowFastaiTimmNFNet-L0ResNeSt-101eMogaNetPandasHugging FaceCNNDeep LearningComputer VisionMedical AI

Graph‑Driven Quant Agents

Intelligent financial advisory system utilizing Model Context Protocol (MCP) and Graph RAG architecture to seamlessly integrate Large Language Models with financial tools. Enables autonomous market analysis, personalized investment strategies, and real-time financial insights through graph-driven knowledge retrieval and agent workflows.

FastAPIPythonPydanticLightRAGOllamaMCPGraphRAGLLMLangChainTypeScriptAI AgentsUvicornAsyncioNomic-EmbedGemma-3-1bFinancial AI

FilmForge LoRA

Custom LoRA (Low-Rank Adaptation) model for Stable Diffusion that generates authentic 35mm film photography aesthetics. Fine-tuned to capture the distinctive grain, color grading, and nostalgic characteristics of analog film photography with high fidelity and artistic quality.

Stable DiffusionLoRAComputer VisionImage GenerationFine-tuningPyTorchDiffusersPhotography AIGenerative AI

EndoSight AI

Collaborative research project with university and private gastroenterology clinic developing AI models for automated detection and classification of gastrointestinal conditions using advanced YOLOv8 object detection combined with U-Net segmentation techniques on endoscopy imaging data.

YOLOv8U-NetObject DetectionSegmentationComputer VisionMedical AI

find more projects on

GitHub
Daniel Cavadia

Thanks for stopping by.
Feel free to reach me through any channel.
I'm excited to connect with you all.

If you have projects and desire to
collaborate with me, I'd love to hear from you.