The Austin ML Journal Club brings together ML/AI practitioners to explore cutting-edge research through focused discussion and collaborative learning. We meet monthly(ish) to dive deep into papers that are shaping our rapidly evolving field. We aim to nurture deep conversations and practitioner insights that reveal how the sausages get made.
All meetings are announced on our LinkedIn page. If you are interested in joining our meetings, please fill out the registration form.
How We Started
The Austin ML Journal Club emerged from a simple recognition: the ML field moves so quickly that staying current with meaningful research is challenging for individual practitioners. While numerous newsletters and articles provide summaries, there’s real value in reading papers deeply, examining methodologies critically, and understanding what actually works in practice.
A journal club creates the structure and accountability for this kind of sustained engagement while transforming what could be isolated study into collaborative learning. What began as a group of Austin-based ML practitioners gathering to critique papers and share workplace insights has evolved to serve a broader community of engineers and researchers who value rigorous discussion over surface-level summaries.
How We Work
Meeting Format
We gather virtually via Zoom for 90-minute sessions designed to fit busy professional schedules. Originally meeting in person in Austin, we transitioned to virtual format to address logistical challenges and serve the growing community of interested participants beyond Austin. Each session centers on a single paper that participants read in advance, allowing for substantive technical discussion rather than surface-level summaries.
Paper Selection
We choose papers that are impactful and seminal in ML/AI, ranging from highly technical research to influential pieces that shape our field’s thinking. Our selections prioritize work that offers both theoretical insights and practical relevance, sparking meaningful debate and learning regardless of technical complexity.
Presentation Format
Sessions are led by volunteer presenters from our community, encouraging diverse perspectives and expertise sharing. The format is entirely up to the presenter - we’ve never required formal slidedecks or PowerPoints, as sharing the paper itself and guiding discussion is sufficient. This approach reduces preparation barriers and keeps our focus on substantive conversation rather than presentation polish. To ensure sustainability and consistency, the organizer serves as a backup presenter when needed, maintaining our regular meeting schedule while fostering community ownership of the learning process.
Discussion Philosophy
We operate under the Chatham House Rule - participants are free to use information shared during meetings, but cannot reveal the source or identity of speakers. This creates a safe space for open intellectual exchange where practitioners can explore ideas, ask questions, and engage in constructive criticism without professional concerns.
Knowledge Sharing
We publish summaries of our discussions on this blog, capturing key insights and diverse perspectives that emerge from our conversations. These summaries serve both as records for participants and resources for the broader ML community. Our blog is built using Quarto and hosted on GitHub Pages, with pull requests automatically published via GitHub Actions, ensuring a streamlined process for sharing our collective insights. See our quarto guideline for blogging tips.
Community Standards
We maintain a welcoming environment for practitioners at all experience levels - from industry engineers to academic researchers to passionate learners. Our community represents diverse industry domains including hardware, software, fintech, and healthcare, as well as academic backgrounds spanning neuroscience, statistics, electrical engineering, astrophysics, biology, and more. This diversity enriches our discussions by bringing varied perspectives to how machine learning intersects with different fields and problem domains. A formal code of conduct is in development to ensure respectful and productive discourse across all backgrounds.
Schedule & Participation
We typically meet late in the month on Thursday evenings at 5:00 PM CT. Registration opens before each session and closes 48 hours prior to maintain focused group dynamics. The club is free and open to anyone interested in advancing their understanding of ML/AI.
Organizer
Hongsup Shin organizes the Austin ML Journal Club, managing meeting coordination, paper selection, and blog maintenance. He is a Principal AI Engineer at Arm with a background in computational neuroscience and behavioral ecology. His interests span AI engineering, MLOps, responsible AI, and AI ethics. Previously, he contributed as a volunteer data scientist at Texas Justice Initiative, a criminal justice non-profit in Austin. He current volunteers as a co-chair of the SciPy conference Proceedings Committee.
Community Contributors
The Austin ML Journal Club has evolved over time with contributions from many dedicated participants. We’re grateful to Joel Afriyie, Brian King, Kshitij Aggarwal, Meghann Agarwal, Saina Lajevardi, Akshata Mohan, Athula Pudhiyidath, and other community members who helped establish the foundations of our collaborative learning environment during the club’s earlier phases.