The Story
Background, philosophy, and how I think about work.
The Story So Far
Building long-term expertise in Machine Learning (research and engineering). Good at communicating and understanding requirements to precise software.
I enjoy prototyping/experimentation to find the true boundaries of what is possible. Taking ownership of a process end-to-end.
Pushing state-of-the-art in Computer Vision has given me confidence to tackle problems where the solution is not obvious. Building a product from zero to one has helped me develop long-term thinking about how base decisions impact the future.
What I Bring
Stay with problems until finding solutions. Aligning myself with team goals quickly. Communicating failures/mistakes early with solutions and follow-up actions.
Natural curiosity to stay near the edge. Mentoring people when necessary.
What I Am Not
Not the 99th percentile in my craft, but made peace with being at 90th and rounding myself in other aspects. Sometimes take smaller mistakes too seriously.
Prefer a good balance between unstructured exploration with defined scope. Do not have experience managing larger teams yet, actively working toward that.
Technical Expertise
Building vision models (2D and 3D) end-to-end: single GPU to multi-GPU training, managing GPU clusters. Establishing MLOps practices for model/dataset reproducibility and traceability.
Key philosophy: developing good ML models means having solid dataset and evals setup. Exploring multi-modal LLMs via prompt engineering and fine-tuning.
Good software engineering practices are accelerators, not slowdowns. Owning the stack from user to compute, end-to-end responsibility for features.