Leonardo Ignacio Córdoba
Verified Expert in Engineering
Software Developer
Buenos Aires, Argentina
Toptal member since March 27, 2019
Leonardo is starting a Ph.D. in CS after working as machine learning engineer in different areas such as transportation, banking, and marketing. He has been involved in the whole process of building models: business problem understanding, data manipulation, EDA, modeling, and deployment. He loves exchanging ideas and he is always looking for cutting edge algorithms and technologies to apply to new problems.
Portfolio
Experience
- Python 3 - 4 years
- Scikit-learn - 2 years
- Pandas - 2 years
Availability
Preferred Environment
Amazon Web Services (AWS), Slack, Jira, Visual Studio Code (VS Code), Git, Linux, Jupyter
The most amazing...
...machine learning model I've built helps banks clients by predicting if they will temporarily need a higher credit card limit, trained with 20 million cases.
Work Experience
Senior Data Scientist
DBI - Havas Group
- Created a scalable serverless system to gather data from many APIs, using AWS stack. This system was made for a big network group in LATAM.
- Built a machine learning model to asses the propensity of a visitor to a web page of buying a subscription. This model uses Google Analytics 360 data and was made for a big Argentinean newsgroup.
- Built a clustering model to make automatic customer segmentation using demographic data and transactional data from a big Colombian company.
- Designed machine learning proposals to tackle business problems in different companies in LATAM.
- Completed exploratory data analysis of clients' databases, communicating insights and visualization of results.
Data Science Professor
Digital House
- Gave lectures on data science theory.
- Taught Python, Pandas, Scikit-learn, GeoPandas, Imblearn, Seaborn, Plotly, etc.
- Developed course materials.
Data Scientist
Prisma Medios de Pago
- Used transactional data (credit and debit cards, ATMs, and POS) a machine learning model for predicting client default was made.
- Created unsupervised machine learning models for clustering analysis were used to generate client segmentation.
- Used geodata different geographic analysis where carried out. For example, market share per region.
- Use transactional data (credit and debit cards, ATMs and POS) a machine learning model for predicting whether a client will exceed its credit limit (and have a transaction denial) was made. This model was used to temporarily raise the credit card limit of a client so as the bank could get more transactions properly done.In addition, the model used a bayesian hyperparameter tuning approach to auto-retrain after some time.
- Deployed models using bash and Control-M. SQL User Store Procedures and bash programmes were used to deploy data manipulation stage in Teradata and in R, prediction stage using models and storing the results back to Teradata.
Data Scientist
Transportation Secretariat - Buenos Aires City Government
- Processed public transportation ticketing system, with about 15 M transactions and 5 M GPS points each day.
- Created automated detection of transfer patterns and of areas a high demand.
- Developed automated bus stops finding using non-supervised machine learning techniques.
- Designed data visualization using R and Carto (geographical data).
- Developed an algorithm for building Origin-Destiny matrix.
- Developed an algorithm for automatically detecting bus lines that don’t fulfill the number of daily services required each day.
- Estimated public and private transport speeds and detecting of traffic problems.
Experience
Medical Appointment No Shows Analysis
https://github.com/LeonardoCordoba/medical_appointment_no_showsEducation
Ph.D. in Computer Science
University of Buenos Aires - Buenos Aires, Argentina
Specialist in Data Science
University of Buenos Aires - Buenos Aires, Argentina
Licenciate (Approximately Equivalent to a Bachelor + Master Degree) in Economics
University of Buenos Aires - Buenos Aires, Argentina
Certifications
Certificate in Advanced English
University of Cambridge
Skills
Libraries/APIs
Scikit-learn, Pandas
Tools
Plotly, GitHub, Jupyter, Git, Jira, Slack
Languages
SQL, Python 3, Python, R
Paradigms
Scrum
Platforms
Linux, Amazon Web Services (AWS), Docker, Visual Studio Code (VS Code)
Storage
PostgreSQL, Microsoft SQL Server, Google Cloud, PostGIS, Teradata, CartoDB
Other
GeoPandas
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