I am an open-source software developer at IBM, where I currently lead the IBM Granite Community on GitHub: maintaining recipes, CI pipelines, and model availability across platforms for every new Granite release.
Recently, I contributed to vLLM and vLLM hardware-enablement plugins for the IBM Spyre Accelerator and maintained KServe/ModelMesh, deepening my expertise in LLM inference and model serving.
Beyond technical contributions, as a developer advocate, I have delivered 10+ talks at international conferences and actively mentor new contributors entering the open-source community.
I hold a MASc with research focused on reinforcement learning and autonomous vehicles, and bring additional experience from the retail data science space. This gives me a well-rounded perspective spanning applied research, production ML systems, and open-source leadership.
Open-Source Contributions
- ibm-granite-community | Maintainer & Project Lead (2026–Present) Updating hands-on workshop recipes for each new Granite model release, maintaining CI build infrastructure, and standing up new models on Replicate and Ollama.
- vLLM-Spyre Plugin | Maintainer (2025–2026) Core contributor and maintainer of the vLLM plugin for the IBM Spyre Accelerator. Focused on test infrastructure, FP8 quantization, sliding window attention, tensor-parallel test suites, CI/CD with model caching, and public documentation.
- vLLM | Contributor (2024–2025) Ranked in the top 12% of ~1,700+ contributors with 15 merged PRs focusing on bug fixes, frontend features, CI linting, and documentation.
- KServe ModelMesh | Maintainer (2022–2024) Maintained a production multi-runtime model-serving platform (Triton, TorchServe, MLServer): dependency/gRPC updates, Grafana observability dashboards, FVT test fixtures, and runtime documentation.
Selected Talks
I frequently speak at international conferences about open-source AI, large language models, and model serving.
- Fast Inference, Furious Scaling — Open Source Summit NA 2025
- Leveraging Open Source AI Securely and Privately — Devopsdays Geneva 2025
- View all talks & workshops →
Publications
- Multi-Objective Autonomous Braking System using Naturalistic Dataset, IEEE Intelligent Transportation Systems Conference, 2019.
- Virtual immersive reality based analysis of behavioural responses in connected and autonomous vehicle environment, Mapping the Travel Behavior Genome, Elsevier, 2020.
- Should I give or should I take? Choice issues in automated vehicle control, International Choice Modelling Conference, 2019.