I am a problem solver, passionate about the power that data gives us to make decisions. With a technical and legal background, (UMINHO) I build my solutions based on data, always trying to understand the details that lead us to error.
I work as data development and analysis in a project related to the area of Finance and Accounting, and I have a special interest in the area of Trading and IA. Love to learn about everything and teach everything I know.
This portfolio addresses different personal projects, more specifically on: 1) Finance & Trading Area, 2) Machine Learning Techniques 3)LawTech (Data & Legal Analytics)
Follow me on social media.
Analytical Tools:
- Data Collect and Storage: SQL, MySQL, Postgres, SQL Server, Impala, Hive, Hadoop Ecosystem, Cassandra, Hue
- Data Processing and Analysis: Python 3, ACL Analytics, Advanced Excel
- Development: Git, Scrum, Linux and Docker.
- Data Vizualization: Power BI, Google Data Studio, Matplotlib, Plotly, Seaborn, Tableau.
- Machine Learning Modeling: Classification, Regression, Clustering, Time Series
- Machine Learning Deployment: Heroku, AWS Cloud, AWS Lambda and Google Cloud Platform ( GCP )
EDA using Techical, Macroeconomic and Sentimental analysis of DXY Index for understand patterns during and after crise like, 2008 (Sub-prime Crisis), Brexit and COVID-19 (2018-2020). Technical steps:
- Python (Pandas |Matplotlib| Seaborn|) for extract data from AlphaVantage using their API, we use price-action fundamentals to recognize patterns of chart behavior, and plot the analysis.
- Web-Scrapping of Trading Economics for understand through the main indicators (e.g GDP) the perspective for the nexts years for the economies related to the DXY index, providing a Macroeconomic analysis.
- We added the social layer through Analyze the sentiment using Spark to collect in time data the Tweets related to DXY posted on StockTwits, using its API for requests. (Tools : Python, Apache Spark,Libraries: PySpark, TextBlob, googletrans, unidecode)
Fraud Detection(Private)
In progress...
- Machine Learning and taking fraud detection to the next level. Companies are reducing their costs with detecting fraudulent transactions, while companies providing theses types of services are increasing thier income. In this project, I built a Machine Learning classifier to label fraudulent transactions with 99.7% of accuracy. The performance of this model would bring revenue of U$898 millions according to the company's business model described in the problem definition.
- Repository: https://github.com/jgoncsilva/Fraud-Detection-BFCompany
Churn Prediction(Private)
*Hey Recruiter ! Ask for ipynb or pdf files i will invite you to see the repo
Python for Financial Analysis, CAPM, Stock Markets Analysis and Visualization, Asset Allocation, Monte Carlo Simulation, Portfolio Optimization, and Trading with Momentum, Bank Market Segmentation
Tasks : Creat and populate an appropriate database, using data collected from web-scrapping from Switchup platafform to create a data model where can dynamically extract data from the past 3 years. We used Matrix BCG for get insights based on position market of Ironhack, for answer the questions
- Whats is IronHack's position regarding the growth of the Coding Bootcamps Market?
- Organic Marketing for evaluate the quality of the Ironhacker community engagement
- How our Market share front our competitors.
Technical Steps
- Build ERD Architecture (python)
- Data Cleaning of the Webscrapped Data( pandas)
- Create a connection between Python and MySQL (PyMySQL Librarie)
- Query for take insights (MySQL Workbench 8.0)) Analytics Consulting Project Tasks : Creat and populate an appropriate database, using data collected from web-scrapping from Switchup platafform to create a data model where can dynamically extract data from the past 3 years. We used Matrix BCG for get insights based on position market of Ironhack, for answer the questions
- Whats is IronHack's position regarding the growth of the Coding Bootcamps Market?
- Organic Marketing for evaluate the quality of the Ironhacker community engagement
- How our Market share front our competitors.
These were my first projects developing in Python. They were very important to develop programming logic. It is very important to know how to build your own functions and documentation. These projects helped me to start with this perspective.
I keep it here also out of affection. Know where I started from.
MySQL |
Python |
pandas |
NumPy |
Matplotlib |
seaborn |
scikit-learn |
SciPy |
statsmodels |
Apache Spark |
Power BI |
Flask |
Heroku |
Streamlit |