Daniel Ramos
Logo Research Engineer

Hello! I'm Daniel, a researcher / engineer at Uber Technologies in the AI Verification and Efficiency team. I have a dual-degree Ph.D. in Computer Science and Software Engineering from Carnegie Mellon University (CMU) and Instituto Superior Técnico (IST).

I research and develop code LLMs and autonomous agents, with interests spanning model training, evaluation, and real-world applications. I’m particularly passionate about foundational models for code and building reliable, efficient AI systems around them.

Google Scholar GitHub Twitter LinkedIn ORCID
Education
  • Carnegie Mellon University
    Carnegie Mellon University
    Aug. 2020 - August 2025
    Ph.D. in Software Engineering
    Thesis: Automated API Migration and Refactoring for Evolving Codebases
    Fully supported by a CMU Portugal Fellowship (~USD 300,000)
  • Instituto Superior Técnico
    Instituto Superior Técnico
    Aug. 2020 - August 2025
    Ph.D. in Computer Science
    Dual Degree Ph.D. program joint with CMU
  • Carnegie Mellon University
    Carnegie Mellon University
    Aug. 2021 - Aug. 2023
    M.Sc. in Software Engineering
    Awarded as a recognition of progress within the Ph.D. program.
  • Instituto Superior Técnico
    Instituto Superior Técnico
    Sep. 2017 - Sep. 2019
    M.Sc. in Computer Science and Engineering
    Thesis: Program Synthesis from Noisy Data
    Machine Learning Track, Grade: 19/20 (first class)
  • Instituto Superior Técnico
    Instituto Superior Técnico
    Sep. 2014 - Sep. 2017
    B.Sc. in Computer Science and Engineering
    Grade: 18/20 (first class)
Experience
  • Uber Technologies Inc.
    Uber Technologies Inc.
    May. 2025 - Present
    Senior Software Engineer, PhD
    Team: AI Verification and Efficiency
    Developer Platform.
  • Uber Technologies Inc.
    Uber Technologies Inc.
    Summer 2023 & 2024
    PhD Research Intern
    Research Group: Programming Systems Group
    Researching and developing a novel polyglot language/toolset (named PolyglotPiranha) for large scale automated refactoring.
    Developed a novel LLM approach for automatically generating tests and other code changes.
  • INESC-ID
    INESC-ID
    Nov. 2019- Aug. 2020
    Junior Researcher
    Research Group: Automated Reasoning and Software Reliability (ARSR)
    Worked on program synthesis for data wrangling and summarization
  • Caixa Magica Software
    Caixa Magica Software
    Summer 2017
    Software Engineer Intern
    Developed an optimizer based on local search to schedule the workers of the largest Portuguese electric utilities company.
Selected Publications (view all )
Are Large Language Models Memorizing Bug Benchmarks?

Daniel Ramos, Claudia Mamede, Kush Jain, Paulo Santos, Catarina Gamboa, Claire Le Goues

2nd International Workshop on Large Language Models for Code (LLM4Code) 2025 Spotlight

Best Paper Award

[Paper]

Are Large Language Models Memorizing Bug Benchmarks?
Are Large Language Models Memorizing Bug Benchmarks?

Daniel Ramos, Claudia Mamede, Kush Jain, Paulo Santos, Catarina Gamboa, Claire Le Goues

2nd International Workshop on Large Language Models for Code (LLM4Code) 2025 Spotlight

Best Paper Award

[Paper]

A Lightweight Polyglot Code Transformation Language

Ameya Ketkar*, Daniel Ramos*, Lazaro Clapp, Raj Barik, Murali Krishna Ramanathan (* equal contribution)

Proceedings of the ACM on Programming Languages (PLDI) 2024 Spotlight

[Paper] [Code]

A Lightweight Polyglot Code Transformation Language
A Lightweight Polyglot Code Transformation Language

Ameya Ketkar*, Daniel Ramos*, Lazaro Clapp, Raj Barik, Murali Krishna Ramanathan (* equal contribution)

Proceedings of the ACM on Programming Languages (PLDI) 2024 Spotlight

[Paper] [Code]

BatFix: Repairing Language Model-based Transpilation

Daniel Ramos, Inês Lynce, Vasco Manquinho, Ruben Martins, Claire Le Goues

ACM Transactions on Software Engineering and Methodology (TOSEM) 2024

[Paper] [Code]

<b>BatFix</b>: Repairing Language Model-based Transpilation
BatFix: Repairing Language Model-based Transpilation

Daniel Ramos, Inês Lynce, Vasco Manquinho, Ruben Martins, Claire Le Goues

ACM Transactions on Software Engineering and Methodology (TOSEM) 2024

[Paper] [Code]

MELT: Mining Effective Lightweight Transformations from Pull Requests

Daniel Ramos, Hailie Mitchell, Inês Lynce, Vasco Manquinho, Ruben Martins, Claire Le Goues

2023 38th IEEE/ACM International Conference on Automated Software Engineering (ASE) 2023 Spotlight

[Paper] [Code]

<b>MELT</b>: Mining Effective Lightweight Transformations from Pull Requests
MELT: Mining Effective Lightweight Transformations from Pull Requests

Daniel Ramos, Hailie Mitchell, Inês Lynce, Vasco Manquinho, Ruben Martins, Claire Le Goues

2023 38th IEEE/ACM International Conference on Automated Software Engineering (ASE) 2023 Spotlight

[Paper] [Code]

All publications