Andre Lustosa

I'm a Principal Software Engineer and Team Lead at Red Hat, working on software foundations, Python packaging, and accelerator enablement. Most of my time now goes into building dependable software, contributing to open source, and leading engineering teams.

I completed my PhD in Computer Science at North Carolina State University in 2025, advised by Tim Menzies. Research is still part of what I do: I occasionally collaborate with Tim on individual projects, but academic research is no longer my main focus. This repository collects selected papers and preserves notes from my PhD work.

For my current engineering work, projects, and writing, visit alustos.us. You can also find my CV, GitHub profile, and Google Scholar profile.

Research

My research has explored how to make software engineering decisions with less data, fewer evaluations, and less effort from the people involved. That includes interactive search, software configuration, hyperparameter optimization, and predicting the health of open-source projects.

Selected publications and preprints

• MOOT: a Repository of Many Multi-Objective Optimization Tasks
  Tim Menzies, Tao Chen, Yulong Ye, Kishan Kumar Ganguly, Amirali Rayegan, Srinath Srinivasan, and Andre Lustosa. MSR 2026, Data and Tool Showcase.
  A shared collection of software engineering optimization tasks for studying trade-offs across competing goals.

• Less Noise, More Signal: the DRR Effect for Better Optimizations of a Range of SE Tasks
  Andre Lustosa and Tim Menzies. IEEE Transactions on Software Engineering, 2026.
  Using the dimensionality reduction ratio to identify when lightweight optimizers can match much costlier methods.

• iSNEAK: Partial Ordering as Heuristics for Model-Based Reasoning in Software Engineering
  Andre Lustosa and Tim Menzies. IEEE Access, 2024.
  Using partial orderings and human feedback to explore software models while reducing the number of questions people need to answer.

• Learning from Very Little Data: On the Value of Landscape Analysis for Predicting Software Project Health
  Andre Lustosa and Tim Menzies. ACM Transactions on Software Engineering and Methodology, 2024.
  Using landscape analysis to tune predictors when project history provides very little training data.

• Streamlining Software Reviews: Efficient Predictive Modeling with Minimal Examples
  Tim Menzies and Andre Lustosa. Preprint, 2024.
  Learning from a small number of expert-labeled examples to guide the exploration of software behavior and configurations.

• Fairer Software Made Easier (using "Keys")
  Tim Menzies, Kewen Peng, and Andre Lustosa. RAISE workshop at ASE, 2021.
  Exploring how a small set of influential features could simplify explanations and fairness analysis.

The titles above link to freely accessible arXiv versions. Entries distinguish published papers from preprints; see Google Scholar for additional publications and citations.

Research archive

These materials capture work in progress from 2022. They are historical notes and reference copies, not a current research plan or publication-status tracker.

• SNEAK 2.0 planning notes: exploratory ideas from February 2022.
• SNEAK: Faster Interactive Search-based SE: the 2021 preprint by Andre Lustosa, Jaydeep Patel, Venkata Sai Teja Malapati, and Tim Menzies, withdrawn in January 2023 for resubmission under a different title. The local manuscript is retained as an archival copy; its folder name reflects the old workflow. For later work on interactive search, see iSNEAK above.
• Background reading retained with those notes: FASTMAP and SWAY.

Last updated: September 2026.