Daniel User

Built AI and data products for merchant acquisition, sales productivity, and growth decision-making at Stone. Worked on predictive scoring systems, AI-assisted sales workflows, and impact measurement products that helped commercial teams prioritize opportunities, evaluate lead quality, and understand whether new initiatives were improving business outcomes.
My work combined product thinking, analy

Worked as Senior Data Scientist II at Jusbrasil.

As a Data Scientist at the leading LegalTech company in LATAM, I drive key AI initiatives that enhance the company’s capabilities. I’ve contributed to pretraining proprietary language models, developing in-house embeddings, and deploying large language models for efficient testing. Additionally, I’ve optimized transformer models for production and developed generative AI solutions for automated do

As a Software Engineer in Globo’s Big Data & AI department, I worked at the intersection of software engineering and machine learning to build and maintain platforms that generate metadata from news media data. My contributions included developing knowledge graphs, implementing audio transcription, assessing brand suitability, classifying customer intent, and annotating news media with taxonomy ca

Ph.D. in Computational Science, focusing on knowledge graph inference using machine learning. My research aimed to create models that predict the truth value of facts within evolving knowledge graphs, with applications in areas like drug repurposing and recommendation systems. This work combines machine learning with real-world challenges to improve decision-making across various fields.

I worked on a research project funded by Petrobras to develop a platform for managing the entire lifecycle of machine learning models, from data preparation to deployment. In addition to conducting academic research, I was involved in various tasks such as defining use cases, data modeling, and implementing prototypes

Master's in computational science with a focus on performance evaluation of graph processing platforms. Beyond planning and executing software performance evaluation, I also analyzed networked data. One of the byproducts of this work was a methodology to suggest which platform was well-suited to a given network dataset (according to features such as scale, clustering coefficient, and diameter) and

Development and performance evaluation of parallel graph algorithms.
Programming languages: C, C++, and Python
Some libraries, frameworks, and APIs: CUDA, MPI, OpenACC, OpenMP, numpy, scipy.

Jul 2022

Nov 2021

May 2019

Jun 2017

Feb 2016

Jul 2016

Feb 2016

Jan 2016