CV
Eduardo Dadalto, PhD in ML
Download PDF
Core Expertise
- AI Evaluation: Metrics design, large-scale evaluation workflows, data processing infrastructure, data lineage and versioning, performance monitoring, sim2real gap computation.
- AI Safety: Conformal risk control, conformal prediction, confidence calibration, uncertainty quantification, robustness testing, out-of-distribution/anomaly detection, operational design domain definition.
- Experimental Design: Bayesian hyperparameter optimisation, statistical hypothesis testing, ablation studies.
Experience
AI Research Engineer
Paris, France | Mar. 2025 -
- Built AI safeguard models end-to-end, deploying them in the HX-2 product, showcasing double digit mission success rate improvements in hard scenarios.
- Developed an ONNX library for edge deployment of conformal prediction and confidence calibration methods.
- Implemented scalable workflows for AI product performance and safety evaluation across the company, ensuring robust deployment and fast R&D iteration cycles.
Co-founder
San Francisco, CA, USA & Paris, France | Apr. 2024 - Nov. 2024
- Pre-trained and fine-tuned transformer models on structured data; released an API platform to customers.
- Raised pre-seed funding from Transpose VC and Entrepreneur First.
Graduate Research Intern
Montreal, QC, Canada | Oct. 2022 - Jan. 2023
Undergraduate Research Intern
Toulouse, France | Apr. 2020 - Oct. 2020
- Trained binary-weight neural networks via projected gradient descent for embedded computer vision.
- Supervised by Franck Mamalet and co-supervised by François Malgouyres and Adrien Gauffriau (Airbus).
Education
PhD in Machine Learning, Computer Science Department
Paris, France | Nov. 2020 - Mar. 2024
You can listen about my thesis here:
(auto-generated with NotebookLM)MSc in Applied Mathematics
Toulouse, France | Sep. 2019 - Oct. 2020
- Dissertation: "Training Binarized Deep Neural Networks." Advisors: Franck Mamalet and François Malgouyres.
Diplôme d'Ingénieur, Major in Data & Decision Sciences
Toulouse, France | Sep. 2018 - Apr. 2020
- Selected Coursework: Supervised, Unsupervised, Deep, and Reinforcement Learning with Emmanuel Rachelson.
- Double Degree with a full tuition fee waiver scholarship.
BE in Aerospace Engineering
São José dos Campos, Brazil | Feb. 2015 - Jul. 2018
- Dissertation: "Machine Learning Applied to Communication Channels." Advisor: Meryem Bennamar.
Publications
For an up-to-date list of publications, please visit my Google Scholar page.
C: Conference · J: Journal · W: Workshop
[C1] Eduardo Dadalto, Marco Romanelli. Optimal Zero-shot Regret Minimization for Selective Classification With Out-of-Distribution Detection. UAI 2025. [link]
[C2] Eduardo Dadalto, Marco Romanelli, Georg Pichler, Pablo Piantanida. A Data-Driven Measure of Relative Uncertainty for Misclassification Detection. ICLR 2024. [ARXIV] [PDF] [CODE]
[J3] Eduardo Dadalto, Florence Alberge, Pierre Duhamel, Pablo Piantanida. Combine and Conquer: A Meta-Analysis on Data Shift and Out-of-Distribution Detection. TMLR 2024. [ARXIV] [PDF] [CODE]
[C4] Maxime Darrin, Guillaume Staerman, Eduardo Dadalto, Jackie CK Cheung, Pablo Piantanida, Pierre Colombo. Unsupervised layer-wise score aggregation for textual OOD detection. AAAI 2023. [ARXIV] [PDF]
[C5] Pierre Colombo, Eduardo Dadalto, Guillaume Staerman, Nathan Noiry, Pablo Piantanida. Beyond Mahalanobis Distance for Textual OOD Detection. NeurIPS 2022. [ARXIV] [PDF]
[C6] Eduardo Dadalto, Florence Alberge, Pierre Duhamel, Pablo Piantanida. Igeood: An Information Geometry Approach to Out-of-Distribution Detection. ICLR 2022. [ARXIV] [PDF] [CODE]
Patents
- [P1] Eduardo Dadalto, M. Donini, G. Detommaso. Method and System for Improving Object Tracking. Filed. 2026.
Preprints
Eduardo Dadalto, Marco Romanelli, Federica Granese, Siddharth Garg, Pablo Piantanida. Trusting the Untrustworthy: A Cautionary Tale on the Pitfalls of Training-based Rejection Option. 2023.
Eduardo Dadalto, Pierre Colombo, Guillaume Staerman, Nathan Noiry, Pablo Piantanida. A Functional Perspective on Multi-Layer Out-of-Distribution Detection. 2023.
Workshops
Eduardo Dadalto, Marco Romanelli, Georg Pichler, Pablo Piantanida. A Data-Driven Measure of Relative Uncertainty for Misclassification Detection. NeurIPS Workshop on Mathematics of Modern Machine Learning. 2023.
Eduardo Dadalto, Florence Alberge, Pierre Duhamel, Pablo Piantanida. Igeood: An Information Geometry Approach to Out-of-Distribution Detection. NeurIPS Workshop on Distribution shifts: connecting methods and applications (DistShift). 2021.
Teaching
CentraleSupélec, Université Paris-Saclay
Teaching Assistant (TA) for the course Introduction to Deep Learning (3SQ2050)
Gif-sur-Yvette, France | 2021-2023
Awards and Scholarships
Nordic Probabilistic AI Summer School (ProbAI)
Norwegian University of Science and Technology
Trondheim, Norway | June 2023
- Sponsored by DeepMind. Coursework on probabilistic modeling, variational inference, and generative models.
Globalink MITACS Scholarship
Awarded for a winning project in a four-month full mobility scholarship program
Montreal, QC, Canada | Oct. 2022 - Feb. 2023
- Partnership between Mila, ÉTS, McGill, Inria, CNRS, and Université Paris-Saclay.
BCG Gamma Data Science Hackathon
Organized by BCG Gamma and Instituto Tecnológico de Aeronáutica
São José dos Campos, Brazil | February 2020
- Fine-tuned a YOLOv3 model for object detection on infrared drone footage.
- Received the Best Presentation Award. Advised by Filipe Verri.
Services & Volunteering
Reviewer: Neural Information Processing Systems (NeurIPS 2022, 2023, ...), Computer Vision and Pattern Recognition (CVPR 2023), International Conference on Learning Representations (ICLR 2024, ...), Conference on Artificial Intelligence (AAAI 2023).
Presentations: Experienced at giving talks to large audiences at academic conferences and industry events.
Teaching: Volunteered as a high school math teacher at a Brazilian public school (2011).
Research Grants
- Compute research grant (AD011012803R) at Jean-Zay, a French HPC/AI cluster. (2021–2024)
- PhD funding from the PSPC AIDA project (2019-PSPC-09) granted by BPI-France. (2020–2024)
Open Source
May 2023
Open source package to accelerate research on out-of-distribution (OOD) detection for computer vision applications.
February 2023
Open source package to accelerate research on out-of-distribution (OOD) detection for textual applications.
Languages: English (fluent), French (fluent), Portuguese (native), German (notions), Spanish (notions).
Programming: Python (PyTorch, Numpy, Tensorflow/Keras, Scikit-Learn, Transformers, MAPIE, Polars, Optuna, etc.), Go, C, CUDA (basics), JavaScript, HTML, SQL, LaTeX, git, Linux, agentic AI tools (Claude Code, Cursor). Large-scale experiments: Slurm, Metaflow, ArgoCD. Edge deployment: Rust (Tokio), ONNX, TensorRT. Cloud: Docker, Terraform IaC, AWS, GCP.
Miscellaneous
- Certificates: Cambridge C2 Proficiency.
- Academic contests: Brazilian Astronomy Olympiad (3× gold medalist); UN mock debating at FAAP/SP.
- Side projects: Co-created Prompt Riddle, an online indie enigma-solving game with 5k+ players.
- Sports: Brazilian jiu-jitsu, futsal.
- Rocketry: Launched and recovered a rocket that reached a 10,000 ft apogee as part of the ITA's Rocket Design team during the 2017 IREC competition (placed 7th/109 teams) in Las Cruces, NM, USA, sponsored by Virgin Galactic.
- Personal interests: Traveling, board games & puzzles.