Building ML systems and distributed inference for edge deployments. Training, tuning, and researching VLMs and LLMs, with applied work on robotic manipulation using deep learning algorithms.
Designed and maintained scalable pipelines processing 2M+ on-chain records. Built automated data quality checks, optimized ETL throughput, and delivered analytics.
Migrated legacy ML workflows to GCP with autoscaling, monitoring, and optimized training/inference pipelines for faster data access and deployment.
Contributed to API definition, debugging, and internal tooling. Improved developer workflows, strengthened documentation, and supported deployment of early ML features.
Worked on vector database integrations, usage guides, and benchmarking. Helped refine workflows around large-scale embeddings and retrieval efficiency.
Conducted applied research on UAV perception and deep vision models. Implemented training pipelines, dataset preprocessing, and evaluation experiments.