Andrea Nalon
Verified Expert in Engineering
Data Scientist and Python Developer
Venice, Metropolitan City of Venice, Italy
Toptal member since March 23, 2016
Andrea is a senior data scientist with extensive experience as a Python programmer and quantitative analyst. He holds a master's degree in engineering and several certifications in quantitative analysis and machine learning. He excels in data analysis, written communication, flexibility, and thinking out of the box. Among Andrea's most notable accomplishments is finding valuable consumer behavior insights for PepsiCo to help guide its marketing campaign across different sub-brands.
Portfolio
Experience
- Data Science - 11 years
- Data Analysis - 10 years
- Python - 8 years
- Algorithmic Trading Analysis - 4 years
- Machine Learning - 4 years
- Algorithmic Trading - 4 years
- Trading Systems - 2 years
- Interactive Brokers API - 1 year
Availability
Preferred Environment
Visual Studio Code (VS Code), Jupyter Notebook, Pandas, Scikit-learn, Python, NumPy, CSV, XLSX File Processing, SQL, Machine Learning
The most amazing...
...thing I've created is a trading strategy for a proprietary trading firm by analyzing trading patterns with statistical and machine-learning techniques.
Work Experience
Financial Data Mining Specialist
Zonal Photon Conversion Inc.
- Handled a very specific request concerning the provision of financial data of the S&P 500 index constituent stocks.
- Suggested relying on Nasdaq Data Link as a data provider and implemented some Python scripts to get all the necessary data over a period of about 26 years.
- Downloaded daily OHLC data for about 1,118 different stocks, as well as market capitalization and price earning ratio.
- Helped the client access a large amount of data by creating several Excel sheets with all the stocks listed as columns and the time series prices on rows.
Day Trading Strategy Expert I Expert Advisor
Private Investor
- Developed and implemented an automated ES-mini futures trading system through the Interactive Brokers API to execute transactions on his account.
- Designed a proprietary algorithm for daily trading with long-term positive returns and built a backtest engine to simulate order execution before live deployment.
- Simulated an annual return of nearly 160% and achieved almost 40% in return in three months of live trading.
- Generated and dispatched daily Excel reports detailing live and backtested trade lists to compare the performance of actual versus backtested operations.
Senior Data Scientist
InvestVerte
- Researched and constructed a backtesting engine tailored for evaluating trading strategies, with a specific focus on the S&P500 trading future ES.
- Used machine learning scikit-learn package to find trading patterns and train effective models, tested with several backtests over many years of historical data.
- Implemented the model using Python and FastAPI, enabling it to operate in real time by connecting with Interactive Brokers to place trading orders. The trading bot could be accessed and queried via the Uvicorn and FastAPI web interface.
Senior Data Scientist
PepsiCo
- Created tools in Databricks and Jupyter Notebook to conduct an in-depth analysis of consumer purchases, identifying correlations and connections among various items, categories, and brands. This helped the marketing sector finalize product campaigns.
- Executed customized scripts to create audience profiles, enabling the marketing team to target product campaigns effectively.
- Increased efficiency in managing large relational datasets with billions of rows using Spark and PySpark, effectively distributing computational tasks across server clusters.
Senior Data Scientist
LGO Capital Holdings
- Applied machine learning techniques to predict cryptocurrency movements for trading.
- Analyzed Bitcoin and Ethereum cryptocurrencies using quotations and additional features from Glassnode.
- Delivered a model to the client leveraging various machine learning tools for classifying the anticipated price movement of the underlying asset, assigning a "1" for a recommended buy the next day and a "-1" for a suggested sell.
- Applied feature selection and scaling throughout the analysis to improve prediction accuracy.
- Employed feature engineering techniques and utilized multiple parallel machine learning models, resulting in an overall system accuracy exceeding 80%.
- Conducted an exhausting test by applying cross-validation methods for backtesting the model from over six years of historical prices.
Data Scientist
Osprey Underwriters
- Created a back-end system and implemented various algorithms to compute the insurance premiums of different products.
- Handled the data cleaning and entire database architecture design across several DB schemas, both manually and via the programming of many Python scripts and Jupyter notebooks.
- Integrated and cooperated on automated tasks running between different servers as required by the customer.
- Customized an already installed and running Moodle server used for video courses–enabling the authentication of users on an external MySQL database with various levels of control.
- Implemented several MySQL stored procedures—using JSON strings as a list of parameters to be transferred to the database—for improved and more accessible integration between various web apps' back and front end.
- Managed the migration of a MySQL production database from one provider (Compose) to another (DigitalOcean) by upgrading its major version from 5 to 8.
- Created a testing MySQL database on Amazon AWS infrastructure, using their RDS service.
- Managed a small team of two developers, coordinating their front-end development work to integrate it with the code and database I created in the back end. I also owned their technical interviews during the hiring process.
Financial Model Builder
Strategic Project Partners
- Migrated all features and elaboration tasks of a complex budget model created with several large Excel workbooks to two Python scripts to accelerate all calculations from over 10 hours to two minutes.
- Established a thorough quality check for input data and implemented constraints to guarantee dependable and resilient execution of the created scripts.
- Created Linux and Windows shell batch programs to run complex data computations automatically.
Quantitative Analyst/Trader
Glory Trading Systems GmbH
- Developed algorithmic trading strategies by analyzing time-series historical data (OHLC and tick data).
- Implemented statistical analysis, linear regression, and machine learning.
- Developed code in R and Python by including markdown documents (knitr and Jupyter notebooks).
Data Analyst
Avepa
- Implemented an automatic report generator to produce statistics reports for the European Commission with detailed payment data.
- Created software to replicate the legacy software used to calculate payments and check the integrity of an internal calculus algorithm.
- Wrote several views and queries to an Oracle back-end database to retrieve payment information.
- Created R scripts to sync a Pentaho repository with an Oracle database to align metadata and descriptions of every Pentaho report.
- Developed a repository with specifications of Oracle views through interviews with different stakeholders.
Data Analyst
City of Treviso
- Created a system for the employment center to statistically analyze labor market data and dynamically create reports with KPI data required by the client.
- Wrote reports with VBA programming to dynamically gather data from the back-end MySQL database.
- Created stored procedures and views in the MySQL back-end database to calculate and filter out unnecessary data.
Business Analyst
GN ReSound
- Created a data warehouse reporting solution to monitor finance, sales, and production departments. The system was connected to the back-end ERP to gather data and create several Excel reports and a Microsoft Access database to interact with.
- Wrote Excel reports with a dynamical update feature where data was downloaded from a back-end database and the cells inside the sheet were filled in and elaborated as required by the client.
- Backed up some data from SQL Server into a local Microsoft Access database for more complex analysis and let the client choose from different filters and sales aggregation; the client could also print reports of his queries.
- Developed software that contained a real-time calendar with upcoming orders and a display of KPI indicators to monitor the production process of hearing instruments.
Experience
A Simple Quantitative Approach of the Three-bar Reversal Pattern
https://nalon99.github.io/publications/three_barsDevelopment of Algorithmic Trading Strategies for a Prop Trading Firm
Machine Learning Applied to Human Activity Recognition
http://nalon99.github.io/Machine_Learning/Music Composition | Mixing and Producing Audio Tracks
https://www.youtube.com/@mrkey-musicEducation
Master's Degree in Computer Engineering
University of Padova - Padova, Italy
Certifications
Unsupervised Learning
Stanford Online University | via Coursera
Advanced Learning Algorithms
Stanford Online University | via Coursera
Supervised Machine Learning: Regression and Classification
Stanford Online University | via Coursera
Cloud Computing Foundations
Duke University | via Coursera
Machine Learning: Classification
University of Washington | via Coursera
An Introduction to Interactive Programming in Python
Rice University | via Coursera
Programming for Everybody: Python
University of Michigan | via Coursera
Practical Machine Learning
Johns Hopkins University | via Coursera
Machine Learning: Foundations
University of Washington | via Coursera
Machine Learning: Regression, Research Methodology, and Quantitative Methods
University of Washington | via Coursera
High Performance Scientific Computing
University of Washington | via Coursera
Introduction to Computational Finance and Financial Econometrics
University of Washington | via Coursera
Financial Markets
Yale University | via Coursera
R Programming
Johns Hopkins University | via Coursera
The Data Scientist's Toolbox
Johns Hopkins University | via Coursera
Getting and Cleaning Data
Johns Hopkins University | via Coursera
Statistical Inference
Johns Hopkins University | via Coursera
Regression Models
Johns Hopkins University | via Coursera
Computational Investing
Georgia Institute of Technology | via Coursera
Mathematical Methods for Quantitative Finance
University of Washington | via Coursera
Skills
Libraries/APIs
NumPy, Pandas, Scikit-learn, SQLAlchemy, Matplotlib, TensorFlow, Keras, Interactive Brokers API, PySpark, REST APIs
Tools
Microsoft Access, Microsoft Excel, Eclipse IDE, Git, GitHub, Moodle, Microsoft Visual Studio, Subversion (SVN), Uvicorn, AWS Fargate, Amazon Elastic Container Service (ECS), Amazon Elastic Container Registry (ECR), Amazon Virtual Private Cloud (VPC)
Languages
SQL, Python, Visual Basic for Applications (VBA), R, Visual Basic, Python 3, XML, Java, Snowflake, Excel VBA, Assembler 68000, Python 2, Markdown, JavaScript
Platforms
Jupyter Notebook, Oracle, Windows, RStudio, Amazon Web Services (AWS), MacOS, Anaconda, Linux, Linux CentOS 7, DigitalOcean, Databricks, Visual Studio Code (VS Code), Docker
Storage
MySQL, PL/SQL, JSON, Oracle PL/SQL, Microsoft SQL Server, SQLite, Amazon DynamoDB
Paradigms
Automation, Anomaly Detection, ETL
Industry Expertise
Trading Systems
Frameworks
Spark
Other
Algorithmic Trading, Data Analysis, Machine Learning, Mathematics, Algorithms, Data Engineering, Back-end, Scientific Data Analysis, Scripting, Data Science, Algorithmic Trading Analysis, Time Series, Time Series Analysis, Supervised Machine Learning, CSV, CSV File Processing, Excel 365, Web Services, Automated Trading Software, Trading Bots, Key Performance Indicators (KPIs), Statistics, Trading, Big Data, Data Mining, Market Basket Analysis, Forex Trading, Bitcoin, Regression, Classification, Statistical Analysis, Financial Data, FastAPI, Multithreading, APIs, Stock Trading, Neural Networks, Bots, XLSX File Processing, Programming, Electronics, CPU Boards, Signal Analysis, Physics, Computer, Networks, Random Forests, Decision Trees, Logistic Regression, Support Vector Machines (SVM), Knitr, Clustering, CI/CD Pipelines, Quantitative Analysis, Data Cleaning, Data Inference, Linear Regression, Linear Algebra, Web Scraping, Audio, Audio Processing, Apple Music, Spotify, Finance, Data, TradingView
How to Work with Toptal
Toptal matches you directly with global industry experts from our network in hours—not weeks or months.
Share your needs
Choose your talent
Start your risk-free talent trial
Top talent is in high demand.
Start hiring