Researcher

Meire Fortunato

Meire Fortunato is a mathematician and AI researcher whose work spans sequence learning, reinforcement learning, mesh-based simulation and global weather forecasting. She earned mathematics degrees at UNICAMP and a doctorate at UC Berkeley, and joined DeepMind as a research scientist in 2016. Her coauthored work includes Pointer Networks, Noisy Networks for Exploration, MeshGraphNets and GraphCast. The GraphCast team won the 2024 MacRobert Award. As of October 2026, she is a Staff Research Scientist at Google DeepMind and a co-founder of Khipu.ai.

Now at
Google DeepMind and Khipu.ai

Timeline

  1. 2007–2010

    Mathematics degrees at UNICAMP

    She earns a bachelor's degree in mathematics at UNICAMP in 2007 and a master's degree there in 2010.

  2. 2015

    Pointer Networks for variable inputs

    In 2015, she coauthors Pointer Networks, which uses neural attention to select elements from a variable-length input sequence and learns approximate solutions to convex hulls, Delaunay triangulations and the planar traveling salesman problem.

  3. 2016

    A Berkeley PhD and DeepMind role

    In 2016, she completes a PhD in mathematics at UC Berkeley, with a thesis on numerical mesh generation, and joins DeepMind as a research scientist in London.

  4. 2018

    Noisy Networks for exploration

    At ICLR 2018, she coauthors Noisy Networks for Exploration, which adds parametric noise to a deep reinforcement learning agent's weights and learns the noise parameters by gradient descent to aid exploration.

  5. 2021

    MeshGraphNets and an ICLR award

    In 2021, she coauthors “Learning mesh-based simulation with Graph Networks,” an ICLR Outstanding Paper. MeshGraphNets uses graph neural networks to learn mesh-based simulations of aerodynamics, structural mechanics and cloth dynamics.

  6. 2023

    GraphCast forecasts global weather

    In 2023, she coauthors GraphCast, a learned weather simulator that uses graph neural networks to produce ten-day global forecasts. The model predicts trajectories for 227 atmospheric variables at six-hour intervals.

  7. 2024

    A MacRobert Award for GraphCast

    The Royal Academy of Engineering awards Google DeepMind the 2024 MacRobert Award for GraphCast; she is a Staff Research Scientist on the winning team.

  8. As of October 2026

    Staff Research Scientist at Google DeepMind

    As of October 2026, she is a Staff Research Scientist at Google DeepMind, and her work centers on reasoning, scientific discovery and agent systems. She is a co-founder of Khipu.ai, which strengthens the Latin American AI ecosystem, and has served as Diversity & Inclusion Co-Chair for WiML at NeurIPS.

Career

  1. Since 2016Google DeepMindResearch Scientist
  2. NowKhipu.aiCo-founder
  3. Women in Machine LearningDiversity & Inclusion Co-Chair

Education

  1. 2016University of California, BerkeleyPhD
  2. 2010University of CampinasBachelor's and master's degrees

Sources and credits

Photo: We have no published photo of Meire Fortunato yet. Send us a free-licensed photo.

  1. Programa de Pós-Graduação em Matemática Aplicada cvlattes.ime.unicamp.br. Checked 21 October 2026.
  2. Pointer Networks proceedings.neurips.cc. Checked 21 October 2026.
  3. Meire Fortunato research.google. Checked 21 October 2026.
  4. Learning mesh-based simulations sites.google.com. Checked 21 October 2026.
  5. Learning Mesh-Based Simulation with Graph Networks arxiv.org. Checked 21 October 2026.
  6. GraphCast: Learned Global Weather Forecasting deepmind.google. Checked 21 October 2026.
  7. 2024 MacRobert Award winner and finalists macrobertaward.raeng.org.uk. Checked 21 October 2026.