WL

Wildson B B Lima

Chemical Engineer | Python Developer | Machine Learning Enthusiast

About Me

As a Chemical Engineer with a Master of Science degree, I possess a strong foundation in the principles of chemical processes and thermodynamics. My expertise extends into the domain of software development, particularly with the Python programming language.

My academic journey and professional endeavors highlight a commitment to leveraging computational tools to advance scientific research. I focus on bridging the gap between scientific research and practical application, translating theoretical concepts into tangible outcomes.

"The most exciting phrase to hear in science, the one that heralds new discoveries, is not 'Eureka!' but 'That's funny...'"

- Isaac Asimov

5+
Research Publications
7+
Years Experience
5+
Open Source Projects

Technical Skills

Programming Languages

Python (PyTorch, Jax, Keras, Django, Kivy)
R for data analysis and statistical computing
JavaScript for web development
Bootstrap, HTML5, and CSS

Python Skills

PyTorch/Machine Learning
Django/Web App
Kivy/App Development
Data Analysis

Frameworks & Libraries

Electron (js)/Kivy (Python) for cross-platform desktop application
ONNX Runtime for optimizing ML models for production
PyTorch Geometric for graph neural networks
Ray for hyperparameter tuning
Weights & Biases (Wandb) for experiment tracking

Cloud & Deployment

Azure, GCP, AWS
Docker and Kubernetes
MLOps workflows
CI/CD practices

Machine Learning

Modeling with scientific research-based methods
Implementing Large Language Models (LLMs)
Graph Neural Networks
Hyperparameter optimization

Featured Projects

GNNPCSAFT Chat

A web application that allows GNNPCSAFT-aware chat interactions with LLMs (Gemini and Ollama), making it possible to ask for thermodynamic properties predictions in chat. Built with Django and deployed as a Docker container and as an Electron app for easy access.

GNNPCSAFT Web App

A web application that provides an intuitive interface for using the GNNPCSAFT model to predict thermodynamic properties. Built with Django and deployed as a Docker container and as an Electron app for easy access.

GNNPCSAFT App

A kivy application that provides another intuitive interface for using the GNNPCSAFT model to predict thermodynamic properties. Built with Kivy and deployed as a desktop application for easy access.

GNNPCSAFT MCP Server

Implementation of the Model Context Protocol (MCP) specifically for GNNPCSAFT tools, such as predicting thermodynamic properties. This server manages communication and context between models and clients using the standardized MCP protocol, facilitating integration with LLM systems.

GNNPCSAFT CLI

A command-line interface for the GNNPCSAFT model, developed so the scientific community can access the model's results easily. Easily installable using pipx.

Education

Publications

Get In Touch