Glenn Matlin
I’m a PhD CS researcher at Georgia Tech advised by Mark Riedl (Associate Director, ML Center) and co-advised by Sudheer Chava (Chair, Finance). I study how to make language-model systems reliable for socio-technical decision-making in high-stakes domains.
My dissertation, RAPID-AI, integrates grounded evaluation, curation-to-SFT-to-RL training, and assurance/interpretability to improve reasoning, analysis, and planning with humans in the loop.
My work has appeared in ACL, COLM, and NeurIPS and has been supported in part by DARPA, Together AI, and collaborators in open-source ecosystems.
Research Focus: RAPID-AI
- Measurement: Grounded, expert-level evaluations of STS knowledge and reasoning from authoritative sources
- Training: SFT and RL methods to improve reasoning, analysis, and planning—not just memorization
- Assurance: Transparency, interpretability, and trust calibration so humans can audit and rely on model recommendations
Experience
Change Healthcare — Healthcare technology and analytics at scale, building ML systems for clinical insights
Komodo Health — Healthcare intelligence platform, real-world evidence generation and analytics
LendUp — Financial inclusion through responsible lending technology and risk modeling
RichRelevance — Personalization and recommendation systems serving millions of users
Technical Interests
LM Systems & Reasoning
Grounded evaluation, SFT/RL for reasoning, interpretability, trust calibration, human-AI interaction
Socio-Technical Domains
Finance, government, security, law, medicine—where outcomes depend on both technical facts and human institutions
Let’s Connect
I’m always interested in discussing research collaborations, interesting technical challenges, and new opportunities. Feel free to reach out via email or connect on LinkedIn.
Explore my research, projects, and blog to learn more about my work.