me

Hi, I’m Bruno Chicelli
Machine Learning Engineer

working at Globo with video, audio and text metadata extraction for digital products. I've also worked with digital pricing optimization and ad-related products in digital and broadcasting environments.

I'm currently transitioning into Graph Neural Networks and improving the current graph applications. I enjoy collaborating and sharing knowledge with nice individuals across the world. Reach out to me if you'd like to connect or discuss potential projects.

Reach out
| | |
Showcase keyboard and mouse image

Education

2020-2022
Polytechnic School of the University of São Paulo
Two year especialization in Artificial Intelligence, focusing in Deep Learning and architecture of Neural Networks.
2015-2019
Federal University of the ABC - São Paulo
Completed 90% of the course. Abandoned this degree in order to join the Specialization in AI.
2015-2018
Federal University of the ABC - São Paulo
Major in Science and Tech.
April 2023 - Current
Staff Machine Learning Engineer - Graph Improvement Squad - Data and AI Dept.
Globo
June 2021 - March 2023
Senior Data Scientist - AdTech Squad - Data and AI Dept.
Globo
August 2020 - May 2021
Data Scientist - Advanced Analytics Squad - Data Science Department
Boa Vista
May 2019 - July 2020
Junior Data Scientist - Innovation in Marketing Squad
Arcelor Mittal

Experience

I like building products in large scale scenarios with data.

Linkedin Download My Resume

Projects

Image to Art Search - WikiArt

Image to Art Search - WikiArt

Find real artwork that looks like any image you want. Made using Vision Transformers w/ Transfer Learning and data from WikiArt and Instagram

Code Demo

3JS Glowing Sphere

Small learning project that I created to teach myself ThreeJS concepts and ThreeJS Fiber (React).

Code Demo
3JS Glowing Sphere
Plotly Calplot

Plotly Calplot

The easiest and best looking Calendar Heatmap you'll find, made with Plotly. I actually got a shoutout from Plotly on Twitter and Linkedin for this one!

Code Medium Article

Todoist Analytics

Keep track and create analytics based on your data from Todoist. If you use Todoist for daily, weekly, monthly or yearly planning you should try this tool. This is how the current version of the tool looks like.

Code Check it out!
Todoist Analytics
NLP Job Classifier (based on linkedin data)

NLP Job Classifier (based on linkedin data)

A Job classification between Data Analyst, Data Scientist and Data Engineer web app made with Tensorflow, Selenium and deployed with Streamlit on Heroku. One simply inputs the job description in the input section and the model trained with ~3k job descriptions scraped with Selenium from jobs in the US.

Code