Work is being done on the implementation of a virtual attendant. For this, NLP techniques were used via the RASA framework. The whole process was deployed in their specific environments, creating images and so on. In this project I also worked on the documentation and scientific research to collaborate with our solutions. Model versioning is also a theme of the project.

In this role, we develop outlier analysis methods through time series analyses. Being developed in python, and collecting data through SQL.

Research on methods of optimization, modeling and spatial analysis in order to allocate families in the state, who need social assistance, in the optimal banks for withdrawing the benefit. Using python libraries like numpy, pandas and one related to operational research, pyomo. I also use data visualization tools such as PowerBI and Tableau, including also for creating dashboards and exploratory analysis.

Main role is to develop intelligence for the company through seasonal reports. I frequently use R and Python for exploratory analysis and PowerBI and/or Tableau for visualization. These analyzes include trend analysis and techniques such as Forecasting and Regression. He also works in the company's research area, where improvements are made to existing tools, with classification and modeling methods.

I worked on an image segmentation project using PyTorch and OpenCV, as well as DeepLearning models using convolutional neural networks.

It was a learning experience, where I came across and improved the main tools in the python language, such as numpy, pandas and ML with the scikit-learn package.