Open to opportunities

Martin Weyssow

@martinweyssow

Message

Research scientist specializing in LLMs, post-training techniques, and production-ready vulnerability detection.

Singapore

Message

What I'm looking for

I seek research-oriented roles that translate LLM research into production, emphasize mentorship, cross-functional collaboration, and opportunities to work on post-training techniques and secure, cost-efficient ML systems.

I am a research scientist focused on large language models, post-training techniques (SFT/RLHF/RLAIF), and their practical deployment for software engineering tasks. My work bridges cutting-edge research and production systems, with emphasis on cost-efficient models and continual learning.

I led a $4M academia–industry collaboration with GovTech Singapore, delivering LLM-based vulnerability detection systems integrated into CI/CD pipelines and outperforming commercial LLMs and static analyzers. I designed R2Vul, an RLAIF approach for robust vulnerability detection.

My research portfolio includes PEFT studies that reduce compute for code generation, CodeUltraFeedback for aligning models to coding preferences, and investigations into LLM-as-a-judge and multi-agent approaches. I secured substantial funding, mentored cross-functional teams, and transitioned research prototypes into production.

I hold a Ph.D. in Computer Science and have a track record of scholarship awards, international collaborations, and teaching. I seek roles where I can continue driving impactful research, mentor engineers, and deliver production-ready ML systems.

Experience

Work history, roles, and key accomplishments

Current

Research Scientist

Current

Singapore Management University

Oct 2024 - Present (11 months)

Led a $4M GovTech Singapore collaboration to develop and deploy LLM-based vulnerability detection systems in CI/CD pipelines, and designed RLAIF-based methods (R2Vul) that outperformed commercial LLMs and static analysers.

Large Language Models RLAIF RLHF PyTorch CI CD Docker

Visiting Researcher

Singapore Management University

Oct 2022 - Jan 2023 (3 months)

Established an international research partnership securing $100,000 NSERC funding and developed parameter-efficient fine-tuning and continual learning techniques to make LLM training cost-effective for code generation.

LoRA)PyTorch Research Collaboration

Education

Degrees, certifications, and relevant coursework

Université de Montréal

Doctor of Philosophy, Computer Science

2020 - 2024

Grade: 4.3/4.3

Activities and societies: Thesis research on continual and preference tuning of LLMs; secured NSERC funding for international collaboration; visiting researcher at Singapore Management University.

Ph.D. in Computer Science focusing on efficient, continual, and preference tuning of LLMs for code generation; awarded with highest distinction.

University of Namur

Master of Science, Computer Science and Data Science

2018 - 2020

Activities and societies: Thesis on code completion using deep learning and static analysis; served as teaching assistant for programming courses.

MSc. in Computer Science and Data Science with a thesis on deep learning and static program analysis for code completion; awarded with highest distinction.