Skip to content
View giulia-balducci's full-sized avatar

Block or report giulia-balducci

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Content in all repositories owned by your account will be closed.
Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. Markdown is supported. This note will only be visible to you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
giulia-balducci/README.md

Hi, I'm Giulia 👋

I'm a Materials Chemistry PhD applying deep scientific expertise to data science, with a focus on climate, energy, and environmental data.

My background is in solid-state chemistry and hydrogen storage materials: air-sensitive synthesis using Schlenk lines and gloveboxes, quantitative Rietveld refinements using GSAS, and neutron diffraction experiments at ISIS Rutherford Appleton Laboratory. During my PhD, I achieved a ~100 K reduction in hydrogen release onset temperature through nanostructuring of LiOH-MgH₂ systems, and proposed and validated dehydrogenation mechanisms for three hydride-hydroxide systems using combined in-situ neutron diffraction, ex-situ PXRD, and thermogravimetric analysis.

What has always driven me is the analytical side: designing experiments, finding patterns in complex data, and communicating results clearly. That's what I do now, with Python instead of lab glassware.


🔬 Background

  • PhD in Materials Chemistry, WestCHEM, University of Glasgow (2011-2015)
  • Master's in Advanced Chemical Methodologies, University of Bologna (110/110 cum laude)
  • Applied AI Programme, AllWomen (2026)
  • 3 peer-reviewed publications in hydrogen storage and materials chemistry
  • 3 patents in materials chemistry and hydroxyapatite electrodeposition
  • Conference presentations at ACS (Dallas) and RSC (St. Andrews, Bath)
  • Experienced science communicator: A-level Chemistry, AS-level Statistics, KS3 Science

🗂️ Projects

Project Description Tools
🪸 Coral Bleaching Risk Prediction End-to-end ML pipeline predicting Atlantic coral bleaching risk from 40 years of oceanographic data. Random Forest optimised for recall (0.82 on test set), deployed on AWS SageMaker with Streamlit dashboard Python, scikit-learn, SHAP, AWS SageMaker, Streamlit
🌊 Marine Heatwave Detection (Mediterranean Sea) SQL/DuckDB pipeline detecting 253,145 marine heatwave events (2016-2026) in the NW Mediterranean using the Hobday et al. methodology, implemented from scratch with window functions and gaps-and-islands logic SQL, DuckDB, Python, xarray
⚗️ Hydride Property Prediction Materials informatics applied to metal hydride systems: Random Forest (R²=0.765) predicts hydrogen storage capacity; honest failure documented for desorption temperature prediction (RMSE=87°C, insufficient data) Python, scikit-learn, SHAP, chemparse
🌿 Greenwashing Detection in Reef-Safe Sunscreen Claims NLP project combining RoBERTa sentiment analysis and VADER on Reddit data, anchored by the ACCC vs. Edgewell regulatory case Python, RoBERTa, VADER, spaCy
🌡️ Land Temperature Trends (1850-2015) Exploratory analysis and regression modelling of global land surface temperatures, asymmetric warming patterns and greenhouse effect signature Python, pandas, seaborn, numpy

🛠️ Skills

libraries  = ["pandas", "numpy", "matplotlib", "seaborn",
              "scikit-learn", "SHAP", "RoBERTa", "VADER", "spaCy"]
ml         = ["random forest", "logistic regression", "SVM",
              "XGBoost", "SMOTE", "clustering"]
deployment = ["AWS SageMaker", "Streamlit"]
domain     = ["hydrogen storage", "materials informatics",
              "climate data", "oceanographic datasets",
              "conservation data"]
tools      = ["Jupyter", "VSCode", "GitHub", "Claude Code"]

🌊 A little more about me

When I'm not wrangling data, I'm probably underwater. I'm a PADI and SSI scuba instructor with 1,600+ dives and 200+ divers certified, around 1,200 of which were in the Coral Triangle. I've lived and worked in the UK, Cambodia, and Malaysia, which means I'm comfortable figuring things out from scratch in unfamiliar environments.


📫 Let's connect

LinkedIn

Popular repositories Loading

  1. coral-bleaching-risk-analysis coral-bleaching-risk-analysis Public

    Coral bleaching risk prediction using 40 years of oceanographic data

    Jupyter Notebook 3

  2. giulia-balducci giulia-balducci Public

    1

  3. land-temperature-trends-1850-2015 land-temperature-trends-1850-2015 Public

    Exploratory analysis and regression modelling of global land temperatures (1850–2015)

    Jupyter Notebook

  4. hydride-property-prediction hydride-property-prediction Public

    Predicting desorption temperature and effective hydrogen release in metal hydride systems to screen candidate materials for near-room-temperature solid-state hydrogen storage.

    Jupyter Notebook

  5. greenwashing-nlp-analysis greenwashing-nlp-analysis Public

    NLP analysis of greenwashing claims in reef-safe sunscreen marketing | VADER · RoBERTa · LDA · spaCy NER | Anchored to ACCC v Edgewell Personal Care (2025)

    Jupyter Notebook

  6. marine-heatwave-detection marine-heatwave-detection Public

    SQL/DuckDB analysis of Mediterranean sea surface temperature (2016-2026) to detect marine heatwave events in the NW Mediterranean, focused on the Tuscan Archipelago.

    HTML