I am a software engineer specializing in deploying and scaling machine learning models in high-volume environments. I work closely with data scientists to ensure models receive accurate, reliable data and operate efficiently at scale. My expertise lies in handling large-scale data pipelines (terabytes per month) and high-throughput systems (10K+ requests per second), optimizing performance and reliability at every iteration.
I have extensive experience across various machine learning fields, including time-series forecasting, anomaly detection, NLP, and reinforcement learning. As a former data scientist, I developed credit scoring models, medical risk prediction systems, and internal frameworks to enhance data science productivity. I also lead over five teams in building high-quality machine learning systems, mentoring engineers, and driving strategic technical decisions.
I hold a Master's degree in Computer Science from the Federal University of Minas Gerais (UFMG), with a focus on Deep Learning, Information Retrieval, and Recommender Systems. I also spent a year studying at Roma Tre University in Italy..
My professional goal is to serve as both a technical and personal mentor, guiding colleagues and newcomers in the field while creating impactful AI-driven solutions.
Tiago Alves
tiagohcalves@gmail.com
Software Engineer• Jan 2022 - Present
CTO• Jul 2021 - Jan 2022
Tech Lead• Jan 2019 - Jul 2021
Data Scientist • July 2016 - Dec 2018
Sofware Engineer • January 2016 - July 2016
Trainee • January 2015 - December 2015
Master in Computer Science • February 2018
Thesis: Dynamic Prediction of ICU Mortality Risk Using Domain Adaptation
B.E. Degree in Computer Engineering • December 2015
Student Exchange • August 2013 to July 2014
This is my very own Data Science course, which i'm writing to get more people interested in the field.
Paper published on the Big Data 2018 Conference
Creative AI is a personal project aimed to give life to machine learning models that help in creativity tasks, such as music and games.
Song Writer is an analsys of lyrics using deep learning. It's also be capable of generating new lyrics from the existing ones.
Quest Generator is a deep learning model to iteractively generate quests for adventure games.
This work developed as a project for the discipline of Data Visualization and intended to visually display information about the story in the books of "Game of Thrones", by George R.R. Martin.
Using a dataset of all pokemon up to the 7th generation, I applied some simple notions of data science to create a pokemon semantic space that allows to find pokemon through math manipulation.