Programmer with 2+ years of experience in data analysis and predictive modeling.
Specialist in Advanced Analytics and AI, focused on process optimization, trend forecasting, and strategic decision-making.
Currently, I work as a Data Scientist in DataQu, where I collaborate with developers from different IT areas.
My main responsibilities cover the entire process: data scraping and analysis, machine learning modeling, and models deployment in web and app environments.
DataQu
The model consists of an ensemble of 5 sub-models. It was trained on data collected by medical personnel, achieving 80% specificity on new data.
It was integrated into the web platform so that doctors can enter patient data and receive real-time predictions.
DataQu
A recommendation algorithm was developed to optimize 200 variables in the mining process (mix extraction, grinder speed, cell temperature, etc.).
A system was created to track and identify rocks throughout the multi-day process.
The model achieved a margin of error of 15 minutes in identification, resulting in a 10% increase in copper recovery.
DataQu
A list of 10,000 accidents was processed, each attributed with 36 possible profiles and 50 potential conditions.
Training was performed using the phi, Llama, and GPT-4 models, both in local and cloud versions.
The model achieved a weighted F1 score of 90% for profile detection and 80% for identifying conditions.
DataQu
An algorithm was created to load and update the real-time Bitcoin price using the Kraken API. An analysis is performed with the most common financial indicators and compared with trend model results.
The results are displayed on an interactive dashboard, which is updated every 15 minutes.