Curriculum Vitae
Founder at Data For Science; previously VP of Data Science & Finance at JPMorgan Chase, Moore-Sloan Data Science Fellow at NYU and faculty at Aix-Marseille Université.
Work Experience
– Present
Founder
Data For Science, Inc., New York NY
- Built a configurable synthetic data generator for insurance claim data using Pyro.
- Created and maintained epidemik, a scientific Python package on epidemic modeling.
- Developed and implemented a pipeline to scrape and parse in real time all on-chain activity for the Stacks blockchain.
- NLP analysis of text data using PyTorch implementation of LSTMs and fine-tuning of LLMs.
- Used LangChain to develop and deploy systems using OpenAI and HuggingFace hosted Large Language Models
- Analyzed social activity leading up to large-scale social movements using geolocated Twitter data and TimescaleDB.
- Built a fuzzy matching approach to match public and private records based on personal information.
- Edited the book Blockchain Data Science, Springer (Forthcoming, 2025).
- Authored blog series on Data Science, Visualization and a Weekly Newsletter.
–
Research Data Scientist
TRM Labs, New York NY
- Developed and implemented Graph based algorithms for tracing and deanonymizing Monero transactions using pySpark.
- Led the implementation and deployment of Provenance and Wash-Trading behavioral signatures for NFTs in Ethereum, Polygon, Binance Smart Chain, and Avalanche.
- Devised and deployed heuristics and algorithms to trace funds between blockchains (Cross-Chain Swaps) for several bridges using Airflow and BigQuery with over 90% accuracy.
–
VP Data Science and Finance
JPMorgan Chase & Co., New York NY
- Analyzed the effects of lockdowns, social distancing measures, and COVID-19 economic recovery using consumer spending in real-time.
- Developed now-casting models of macro-economic variables by using Hive and pySpark to leverage tens of billions of card transactions.
- Quantified market influence by NLP and sentiment analysis applied to public Federal Reserve statements.
- Modeled customer response to rate changes by applying Causal Inference to advise internal rate policy changes.
–
Moore-Sloan Data Science Fellow
New York University, New York NY
- Developed and implemented a global event detector using Hadoop to process Wikipedia MySQL dumps and real-time Wikipedia page view data.
- Defined the boundaries of linguistic dialects in large-scale Twitter data by applying PCA and KMeans.
–
Maître de Conférences
Aix-Marseille Université, Marseille, France
- Modeled human mobility patterns between cities to understand tourism travel using a large-scale geocoded Twitter dataset.
- Developed a Multiplex (multi-layer) network agent-based model for the influence of human mobility in social structure.
- Edited a book on Social Phenomena: From Data To Models, Springer (2015).
–
Associate Research Scientist
Northeastern University, Boston, MA
- Predicted (in advance) the winner of the 11th edition of American Idol using a stream of real-time geolocated Twitter data as a case study of using online platforms for the forecasting of real-world events.
- Investigated how the Global Language Network impacts cultural diffusion using NLP and Network Science techniques to analyze multilingual Twitter, Wikipedia and UNESCO datasets.
–
Research Associate
Indiana University, Bloomington, IN
- Developed and implemented the Global Epidemic and Mobility Modeler (GLEaM) system for large-scale modeling of epidemic spreading.
- Detected the influence of automated bots in online political discussion using Applied Machine Learning methods (SVM, LSA, and AdaBoost).
Education
–
Ph.D. in Physics
Emory University
–
M.S. in Computer Science
Emory University
–
M.S. in Physics
Universidade de Aveiro
–
B.S. in Physics
Universidade do Porto