Summary
Overview
Education
Work History
Scientific Research
Certification
Timeline
Attilio Sbrana

Attilio Sbrana

New York

Summary

Vice President of Quantitative Strategies at Goldman Sachs with over 12 years of experience driving innovation in financial modeling, risk analysis, and AI/ML applications. Recognized for developing cutting-edge deep learning and machine learning solutions that have optimized investment strategies, enhanced risk management, and improved operational efficiency. Holds a Ph.D. in Electronic and Computer Engineering with a focus on Transformer models and has made significant scientific contributions to the field of AI/ML, with publications in top conferences and journals. Brings a unique blend of technical expertise, strategic thinking, and leadership to deliver transformative results in complex financial environments.

Overview

12
12

Years of professional experience

6
6

Published papers

Education

Doctor of Science in Electronic and Computer Engineering -

Aeronautics Institute of Technology (ITA), São José dos Campos, Brazil
2023

GPA: 9.9/10

Master of Science in Computer Science -

Federal University of São Carlos (UFSCar), São Carlos, Brazil
2020

GPA: 4.0/4

Master of Business Administration -

Institut Européen D'Administration Des Affaires (INSEAD), Fontainebleau, France
2018

GPA: 3.3/4

Bachelor of Science in Economics -

Insper Instituto De Ensino E Pesquisa, São Paulo, Brazil
2010

GPA: 8.1/10

Work History

Vice President, Quantitative Strategies

Goldman Sachs
01.2024 - Current
  • Financial Models: Spearheading the development and enhancement of quantitative financial models, leveraging advanced statistical and econometric methodologies.
  • AI Initiatives: Leading initiatives in Applied AI, focusing on machine learning, deep learning, natural language processing, and time series analysis to solve complex financial challenges.
  • AI/ML Systems: Designing, testing, and maintaining sophisticated AI/ML-based software systems for enhanced financial decision-making and strategic planning.
  • Risk Management: Implementing robust risk management frameworks and ensuring adherence to regulatory compliance standards in wholesale risk modeling.

Principal Data Scientist, Artificial Intelligence

Wedbush Securities / Qapital
05.2022 - 12.2023
  • LLM Development: Played a pivotal role in the development of Large Language Models (LLM) at Wedbush Securities, providing comprehensive support across all business areas.
  • Marketing Optimization: Lead a team of four data scientists to develop intelligent systems for marketing optimization. Leveraged control heuristics and machine learning modeling, leading to a decrease in customer acquisition costs by over 50%. Additionally, product A/B testing strategies resulted in a notable increase in customer retention, up by more than 20 percentage points.
  • Customer Segmentation: Developed customer segmentation models for predicting customer lifetime value and crafted deep learning models to analyze customer behavior. These models generated valuable insights, crucial in shaping new marketing campaign strategies. These actions resulted in a 50% decrease in net churn, and a 30% increase in expected LTV.

Research Scientist, Artificial Intelligence

Flex
02.2021 - 04.2022
  • AI Solutions: Led a six-person R&D team, pioneering advanced AI solutions organization-wide. Designed and executed an extensive server infrastructure, powered by eight A100 GPUs and orchestrated by Kubernetes, slashing model training and deployment time by a factor of five.
  • Computer Vision: Led a computer vision project in plastic recycling, developing an AI system to identify black-polymers via near infra-red hyperspectroscopy. The work, published in the IEEE Sensor Letters, boosted recycling yields with over 70% accuracy for black plastic separation and over 90% for colored plastic.
  • Anomaly Detection: Assisted in creating an automated machine learning audio anomaly detector for Motorola's cellphone line, achieving over 90% defect identification and accurate causality determination, enhancing quality control and repair precision.
  • Knowledge Graph: Developed a knowledge-graph database and visualization solution for Cisco using Gremlin language and JanusGraph, streamlining their data analysis process and enabling informed decision-making.
  • Forecasting Models: Made significant contributions to developing precise forecasting models that refined the company's production planning.

Graduate Research Scientist, Artificial Intelligence

Aeronautics Institute of Technology (ITA)
01.2020 - 01.2021
  • Optimization Models: Developed optimization models for large-scale combinatorial optimization problems, leading to a deep learning agent capable of 98% accuracy in NP-Complete Akari games, a feat documented in the 'Solving NP-Complete Akari Games with Deep Learning' paper, published in the journal Entertainment Computing.
  • Time-Series Forecasting: Spearheaded research in deep learning and time-series forecasting, creating financial forecasting methods with an emphasis on cryptocurrency.
  • Submitted 'N-BEATS Perceiver: A Novel Approach for Robust Cryptocurrency Portfolio Forecasting' for review in Computational Economics and created a comprehensive database of cryptocurrency price data.
  • NLP Research: Co-authored the innovative NLP paper 'Developing and Assessing a Human-Understandable Metric for Evaluating Local Interpretable Model-Agnostic Explanations,' enhancing 'black-box' deep learning model interpretability. Published in the Journal of Intelligent Engineering and Systems.
  • Research Grant: Co-recipient of a Research Grant to finance the above-mentioned research endeavors.

Graduate Research Scientist, Artificial Intelligence

Federal University of Sao Carlos (UFSCar)
01.2019 - 12.2019
  • Optimization & Forecasting: Conducted two parallel research projects in the fields of heuristic optimization, computational mathematics, and deep learning for time-series forecasting, resulting in the publications 'N-BEATS-RNN: Deep Learning for Time Series Forecasting' presented at the IEEE International Conference on Machine Learning and Applications and 'GPU Accelerated Metaheuristics for Integrated Production Lot Sizing and Scheduling Problems' presented at the IEEE International Conference on Artificial Intelligence in Engineering and Technology. Also created a software package VectOpt for metaheuristic optimization of lot sizing and scheduling problems and portfolio optimization.
  • Teaching Assistant: Served as teaching assistant for the Big Data and Massive Parallel Processing course at UFSCar. Developed and delivered tutorials on MongoDB, Neo4J and Redis for students. Awarded a merit-based scholarship for research and co-beneficiary of a research grant.
  • Churn Prediction: Researched and developed a user churn prediction machine learning algorithm in collaboration with a private gaming company, Gamers Club, attaining nearly 90% accuracy in forecasting player churn aligned with engagement patterns. Concurrently, contributed to the development of anomaly detection for preventing fraud in online multiplayer games by examining in-game performance.

Quantitative Researcher, Commodities

Pala Investments
06.2018 - 10.2018
  • Investment Strategies: Implemented investment strategies, utilizing advanced methods such as neural networks, decision trees, evolutionary algorithms, and portfolio construction algorithms, resulting in a decrease in the portfolio's risk profile.
  • Microeconomic Models: Constructed microeconomic models estimating supply and demand for uranium, cobalt, and nickel using regression and linear programming. These models equipped the fund with insights to formulate informed investment decisions and strategize suitable investment approaches.
  • Risk Assessments: Performed comprehensive geological risk assessments for three distinct Private Equity investments in the nickel and copper sectors, of nearly $200 million.

Executive Director, Investment Banking

Wimmer Financial
10.2016 - 11.2017
  • Team Management: Managed a team of 5 to 6 financial analysts and engineers while increasing production from 1 deal per quarter to 4 deals per quarter as the Executive Director of Wimmer Financial, an advisory boutique specializing in financial structuring for commodity investments.
  • Structured Transactions: Structured quantitative transactions for commodity assets globally, utilizing expertise in structured finance, derivatives, and statistical risk management, to drive firm performance by $2 million yearly.
  • Complex Projects: Engaged in complex projects encompassing commodity statistics, geological resource estimates, commodities production planning, financial planning, prepays, streaming contracts, restructuring, and mergers and acquisitions. Delivered strategic solutions to meet client needs, completing over six risk modeling projects for entire mining operations within a year, triggering a 5x increase in company profits.

Senior Associate, Investment Banking

Goldman Sachs
03.2013 - 05.2016
  • Structured Transactions: As a member of the Financing Group within the Investment Banking Division, provided structured transaction expertise to important clients in the Americas, including corporations, financial institutions, and sovereigns.
  • Client Portfolio: Managed a client portfolio of financial institutions, generating about $12mm in revenues per year to the firm with medium term notes programes and associated interest rate derivatives.
  • Capital Instruments: Structured and priced $2 billion of complex capital instruments as total return swaps, generating $200 million of net revenues.
  • Team Leadership: Oversaw a group of five to six financial analysts, accruing substantial project management and team leadership experience Responsibilities included daily task allocation and provision of informal feedback.

Senior Analyst, Credit Risk

Goldman Sachs
08.2010 - 02.2013
  • Risk Exposure Reduction: Leveraged financial engineering and applied statistics expertise to minimize credit risk exposure, saving over $50 million through stress testing and statistical analysis of derivatives and credit operations in two Brazilian bank failures.
  • Structured Loans: Orchestrated 25+ structured loans across North American mid-cap companies in oil & gas, energy, and mining sectors, managing 250+ counterparties in North America and Latin America. Provided valuable client advisement on ratings and the impact of corporate actions.
  • Leadership Recognition: Acknowledged leadership and expertise through promotions and committee appointments, including Secretary for the Natural Resources, Latin America, and Credit Analytics Committees.

Analyst, Market & Liquidity Risk

Itau
05.2010 - 07.2010
  • Exposure Analysis: Conducted comprehensive analysis of interest rate and currency exposure in Latin America, and assessed against the Euro via the Portuguese branch of the bank.
  • Financial Metrics: Developed an efficient codebase to calculate critical financial metrics, including VaR, Stress Test, and DV01.
  • Regulatory Compliance: Assured compliance with Central Bank regulations and delivered accurate internal reports by meticulously maintaining and updating the aforementioned codebase.

Scientific Research

Conference Articles as Primary Author

  • Sbrana, A., Debiaso Rossi, A. L., & Coelho Naldi, M. (2020). "N-BEATS-RNN: Deep Learning for Time Series Forecasting," in Proceedings of the 19th IEEE International Conference on Machine Learning and Applications (ICMLA, Miami, FL, USA, pp. 765-768. https://doi.org/10.1109/ICMLA51294.2020.00125.
  • Sbrana, A., Ferreira, D., & Cantão, R. F. (2022). "GPU Accelerated Metaheuristics for Integrated Production Lot Sizing and Scheduling Problems," in Proceedings of the IEEE International Conference on Artificial Intelligence in Engineering and Technology (IICAIET), Kota Kinabalu, Malaysia, pp. 1-6. https://doi.org/10.1109/IICAIET55139.2022.9936782.


Journal Articles as Primary Author

  • Sbrana, A., de Almeida, A., de Oliveira, A. G., Neto, H. S., Rimes, J. P. C., & Belli, M. C. (2023). "Plastic Classification With NIR Hyperspectral Images and Deep Learning," IEEE Sensors Letters, vol. 7, no. 1, Art no. 6000404, pp. 1-4. https://doi.org/10.1109/LSENS.2023.3234401.
  • Sbrana, A., Mirisola, L. G. B., Soma, N. Y., & de Castro, P. A. L. (2023). "Solving NP-Complete Akari Games with Deep Learning," Entertainment Computing, vol. 47, 100580. https://doi.org/10.1016/j.entcom.2023.100580.
  • Sbrana, A., & de Castro, P. A. L. (2023). "N-BEATS Perceiver: A Novel Approach for Robust Cryptocurrency Portfolio Forecasting," Computational Economics. https://doi.org/10.1007/s10614-023-10470-8.


Co-authored Articles

  • Silva, R. J. O., Sbrana, A., de Castro, P. A. L., & Soma, N. Y. (2023). "Developing and Assessing a Human-Understandable Metric for Evaluating Local Interpretable Model-Agnostic Explanations," International Journal of Intelligent Engineering and Systems (IJIES). https://doi.org/10.22266/ijies2023.0831.26.

Certification

  • IEEE Senior Member, 2023, Institute of Electrical and Electronics Engineers, Positions among the top 10% of over 400,000 IEEE members worldwide, emphasizing exceptional professional recognition.
  • Certified Peer Reviewer, 2022, Elsevier, Enabled execution of 40+ peer reviews, enhancing the quality of scholarly communication.

Timeline

Vice President, Quantitative Strategies - Goldman Sachs
01.2024 - Current
Principal Data Scientist, Artificial Intelligence - Wedbush Securities / Qapital
05.2022 - 12.2023
Research Scientist, Artificial Intelligence - Flex
02.2021 - 04.2022
Graduate Research Scientist, Artificial Intelligence - Aeronautics Institute of Technology (ITA)
01.2020 - 01.2021
Graduate Research Scientist, Artificial Intelligence - Federal University of Sao Carlos (UFSCar)
01.2019 - 12.2019
Quantitative Researcher, Commodities - Pala Investments
06.2018 - 10.2018
Executive Director, Investment Banking - Wimmer Financial
10.2016 - 11.2017
Senior Associate, Investment Banking - Goldman Sachs
03.2013 - 05.2016
Senior Analyst, Credit Risk - Goldman Sachs
08.2010 - 02.2013
Analyst, Market & Liquidity Risk - Itau
05.2010 - 07.2010
Aeronautics Institute of Technology (ITA) - Doctor of Science in Electronic and Computer Engineering,
Federal University of São Carlos (UFSCar) - Master of Science in Computer Science,
Institut Européen D'Administration Des Affaires (INSEAD) - Master of Business Administration,
Insper Instituto De Ensino E Pesquisa - Bachelor of Science in Economics,
Attilio Sbrana