Mariana Clare
Machine Learning And Uncertainty Quantification Scientist (Data-Driven Weather Models)
Bonn, North Rhine-Westphalia, Germany
Summary
👤
Senior🎓
Top SchoolMariana Clare is a machine learning and uncertainty quantification scientist with eight years’ experience applying probabilistic AI to weather and climate systems, currently helping develop ECMWF’s data-driven forecasting model AIFS. She bridges rigorous mathematical modelling (PhD, Imperial College) with operational forecasting, focusing on trustworthy, interpretable methods—teaching ECMWF’s MOOC that reached 7,000+ learners. Her work spans generative and Bayesian neural networks for probabilistic forecasts, adjoint-enabled finite-element contributions to Firedrake, and practical post-processing for high-resolution digital twins. Prior roles at the Met Office, IPSL and industry (PwC, MUFG) illustrate her ability to translate research into deployable tools and stakeholder-facing products. Notably, she pairs deep domain expertise in coastal and ocean dynamics with hands-on open-source commits that improve uncertainty workflows in community projects.
8 years of coding experience
3 years of employment as a software developer
MMath Mathematics, MMath Mathematics at University of OxfordPhD Mathematical modelling Department of Earth Science and Engineering, PhD Mathematical modelling Department of Earth Science and Engineering at Imperial College LondonSouth Hampstead High School, GDST
Github Skills (20)
python10
firedac10
machine-learning10
adx10
el10
adt10
adp10
f10
tensorflow10
finite-element-analysis10
jupyter-notebook10
forms9
data-analysis9
compile8
compilation8
Programming languages (2)
Github contributions (5)
firedrakeproject/firedrake
Jan 2020 - Jan 2022
Firedrake is an automated system for the portable solution of partial differential equations using the finite element method (FEM)
Role in this project:
Back-end Developer
Back-end DeveloperContributions:5 reviews, 27 commits, 1 PR in 1 year 11 months
Contributions summary:Mariana focused on enhancing the adjoint capabilities of the Firedrake project. They added new functionality to split functions and modified existing blocks, constants and function mixins to support adjoint calculations. The contributions involved modifications to core components and also refactoring some aspects of the existing adjoint system. The user appears to be improving the performance and accuracy of adjoint operations within the finite element method framework.
equationspartialfemmethodsimulation
ecmwf-projects/mooc-machine-learning-weather-climate
Feb 2023 - Mar 2023
Role in this project:Data Scientist
Data ScientistContributions:6 commits, 19 pushes in 12 days
Contributions summary:Mariana primarily focused on updating a Jupyter Notebook related to probabilistic ocean regime predictions using a Bayesian Neural Network (BNN). Their commits involved modifying code within the notebook, specifically in the `tier_2/uncertainty/Using a BNN for probabilistic ocean regime predictions.ipynb` file, likely related to model compilation and summary. They also updated `obs-env.ipynb` related to observations and the test file `test_notebooks.py`.
Find and Hire Top DevelopersWe’ve analyzed the programming source code of over 60 million software developers on GitHub and scored them by 50,000 skills. Sign-up on Prog,AI to search for software developers.