+++ title = "Experience" +++
@def mintoclevel=1
\toc
<style>
h1 {
font-size: 24px;
margin-top: 0;
}
h2 {
font-size: 22px;
}
h3 {
font-size: 18px;
}
p {
font-size: 18px;
line-height: 1.4;
}
.section {
margin-bottom: 30px;
}
.sub-section {
margin-top: 10px;
}
.sub-section p {
margin-left: 20px;
}
</style>
<div class="section">
<div class="sub-section">
<h2>2023-Present: PhD in Machine Learning and Optimization</h2>
<p><em>H. Milton Stewart School of Industrial and Systems Engineering, Georgia Institute of Technology, Atlanta, US</em></p>
<p><strong>Advisor:</strong> <a href="https://www.isye.gatech.edu/users/pascal-van-hentenryck">Dr. Pascal Van Hentenryck</a></p>
<p><strong>Research:</strong> Decision-making methods under uncertainty, combining Machine Learning and Optimization. Focus on Energy Systems and Sustainability.</p>
</div>
<div class="sub-section">
<h2>2017-2019: Master's degree in Operations Research</h2>
<p><em>Electrical Engineering Department, Pontifical Catholic University of Rio de Janeiro, PUC-Rio, Brazil</em></p>
<p><strong>Advisor:</strong> <a href="https://alexandrestreet.com/">Dr. Alexandre Street</a></p>
<p><strong>Committee:</strong> <a href="https://scholar.google.co.uk/citations?user=AJ6YFH4AAAAJ&hl=en">Andy Philpott</a>, <a href="https://www.sintef.no/en/all-employees/employee/arild.helseth/">Arild Helseth</a>, <a href="http://www.ind.puc-rio.br/en/equipe/davi-michel-valladao/">Davi Valladão</a>, and <a href="http://www.ind.puc-rio.br/en/equipe/bruno-fanzeres-dos-santos/">Bruno Fanzeres</a></p>
<p><strong>Focus:</strong> Optimization and Decision Under Uncertainty approaches for quantitative finance; portfolio optimization; dynamic allocation of resources; asset and liability management.</p>
<p><strong>Significant Classes:</strong> Convex and Integer programming, decision under uncertainty (covering Online Learning, Robust and Stochastic Optimization), and Probability Theory (covering high dimensional statistics, structured and score-driven models for forecasting, and measure theory).</p>
<p><strong>Thesis:</strong> A Framework for Assessing the Impacts of Network Formulations in the Operation of Hydrothermal Power Systems - DOI: <a href="https://doi.org/10.17771/PUCRio.acad.51577">10.17771/PUCRio.acad.51577</a></p>
</div>
<div class="sub-section">
<h2>2014-2017: Double Degree, General Engineering, BRAFITEC Merit Scholarship</h2>
<p><em>École Centrale de Marseille, France</em></p>
<p><strong>Minor Degree:</strong> Computer Science and Information Systems.</p>
<p><strong>Significant Classes:</strong> Artificial Intelligence (supervised and unsupervised learning for regression and classification), Information Theory, Optimal Control, and Continuous Analysis.</p>
</div>
<div class="sub-section">
<h2>2012-2017: Bachelor's degree in Mechatronics and Control Engineering</h2>
<p><em>Pontifical Catholic University of Rio de Janeiro, PUC-Rio, Brazil</em></p>
<p><strong>Significant Classes:</strong> Linear Programming, Functional and Object-oriented programming, Data Structures, and Mathematical Programming in Capital markets.</p>
</div>
</div>
<div class="section">
<div class="sub-section">
<h2>2023-Present: Researcher (Machine Learning and Optimization)</h2>
<p><em>AI4OPT, Atlanta, US</em></p>
<p>Research in automated decision-making methods under uncertainty, combining Machine Learning and Optimization. The main emphasis is on the development of interpretable AI for pressing global challenges using tractable models and robust decision frameworks.</p>
</div>
<div class="sub-section">
<h2>2020-2023: Researcher (Machine Learning and Optimization)</h2>
<p><em>Invenia Labs, Cambridge, UK</em></p>
<p>Strategic algorithmic operation in energy markets to support the renewable energy transition.</p>
<ul>
<li>Led projects on the implementation of stochastic, robust, and distributionally robust optimization approaches to tackle uncertainty in daily bidding operations, coupling Machine Learning techniques to estimate uncertainty and appropriate optimization methods to find robust decisions.</li>
<li>Contributed to the construction of a detailed market simulator for counterfactual analysis. Led the data analysis of simulation results accuracy.</li>
<li>Led work on decomposition techniques for large-scale Integer and continuous optimization problems, proposed methods for estimation of unknown parameters of the grid through Inverse Optimization and other heuristics.</li>
<li>Exploration of exact and surrogate approaches to bi-level optimization (for hyperparameter optimization and strategic bidding). Including integration of Automatic Differentiation and Automatic Dualization of optimization problems to help implement explored approaches.</li>
<li>Experience in MISO, PJM, ERCOT, CAISO, SPP, ISO-NE, and NYISO.</li>
</ul>
</div>
<div class="sub-section">
<h2>2017-2019: Research Staff Member</h2>
<p><em>LAMPS PUC-Rio, Rio de Janeiro, Brazil</em></p>
<p>Laboratory for Applied Mathematical Programming and Statistics (LAMPS).</p>
<ul>
<li>Research and development of mathematical programming solutions to find power systems transmission usage optimal contracting strategies for one of Brazil's biggest electricity services companies (Energisa Group).</li>
<li>Development of flexible software for Hydrothermal Multistage Economic dispatch in a consulting project for FGV Energia to help assess network and installed capacity changes. (HydroPowerModels Project).</li>
<!-- Add other points as needed -->
</ul>
</div>
<div class="sub-section">
<h2>2017: Intern</h2>
<p><em>Bank BBM, BoCom, Rio de Janeiro, Brazil</em></p>
<p>BOCOM BBM provides credit and financial services to corporate clients in the Corporate and Large Corporate sectors.</p>
<ul>
<li>Trade-note receivable fraud-risk assessment while in the Data Science team.</li>
<li>Macroeconomic indicators analysis while in the Economic Research team.</li>
<!-- Add other points as needed -->
</ul>
</div>
<div class="sub-section">
<h2>2016-2017: Research Intern</h2>
<p><em>Air Liquide Medical Systems, Air Liquide Group, Paris, France</em></p>
<p>Modeling of the company's first integrated breathing-support machine simulator to help researchers test new designs and access solutions. Focus on both CPAP and BiPAP breathing programs for the Vendome home ventilator.</p>
</div>
<div class="sub-section">
<h2>2015: Intern</h2>
<p><em>Northumberland Tyne and Wear NHS Foundation Trust, Newcastle, UK</em></p>
<p>Creation of a digital consumables database for stock control and protocols for logging supply usage.</p>
</div>
</div>