Joseph Lattanzi

Hello! Glad to see you here. This is a collection of my data science projects focused on machine learning, optimization, and analytics.

ICU Readmission Prediction

Who Comes Back? ICU Readmission Prediction

Machine Learning | R | XGBoost | MIMIC-IV

End-to-end ML system predicting 30-day ICU readmissions with 0.683 AUC, delivering $17.25M potential annual value.
NFL Survivor Pool Optimizer

NFL Survivor Pool Optimizer

Python | Streamlit | Optimization | Hungarian Algorithm

Data-driven optimization system using the Hungarian algorithm to maximize survival probability in NFL pools.
Box Box F1 Fantasy

Box Box: F1 Fantasy Head-to-Head

Next.js | TypeScript | Prisma | PostgreSQL

Full-stack F1 fantasy game with real-time scoring from live race data. Pick drivers, constructors, and bet on race chaos with friends.
SEC Risk Factor Intelligence

SEC 10-K Risk Factor Intelligence

Python | NLP | BERTopic | SHAP | Transformers

End-to-end NLP pipeline analyzing 79,000+ SEC risk disclosures with topic modeling, classification (71.2% F1), and semantic similarity analysis.
RAG Research Assistant

RAG Research Assistant

Python | LangChain | Llama 3.1 | ChromaDB | Streamlit

Production-grade Retrieval-Augmented Generation system with semantic search, vector storage, and LLM-powered Q&A with source citations.
PokeForge LoRA Fine-Tuning

PokeForge: Fine-Tuning Diffusion Models with LoRA

PyTorch | Stable Diffusion | LoRA/PEFT | Deep Learning

Parameter-efficient fine-tuning of Stable Diffusion using LoRA adapters, achieving 95% reduction in trainable parameters on consumer hardware.
Joseph Lattanzi

About Me

I'm a data scientist who enjoys solving problems with data. I build machine learning models, optimize algorithms, and create tools that help make better decisions.

My work spans the full pipeline—from cleaning messy datasets to deploying models and applications that people can actually use. I focus on understanding the problem first, then finding the right technical approach to solve it.

Technical skills:

  • Languages: Python, R, SQL
  • ML & Statistics: XGBoost, Random Forest, Deep Learning, NLP, Statistical Modeling
  • End-to-End: Data extraction, feature engineering, model development, deployment

I'm always learning new techniques and looking for interesting problems to work on. Feel free to reach out if you'd like to connect.

Get In Touch

Feel free to reach out for collaborations or just a friendly chat.

lattanzi.joseph@gmail.com
linkedin.com/in/joseph-lattanzi
github.com/jlattanzi4