I've built production AI systems for Eyecare, Pix Force, and Hoobox, taking computer vision, document AI, and generative AI workflows from research through deployment.
At Eyecare, I design and operate ophthalmology inference services, coordinating vision and speech models for autorefractor exams under strict latency budgets. Previously at Pix Force, I deployed PyTorch and YOLOv8 models for industrial safety and document extraction, supported by FastAPI, RabbitMQ, Docker, and AWS.
I build across the full ML lifecycle: data pipelines, training, evaluation, deployment, monitoring, and retraining. My work connects clinical, product, and engineering stakeholders through reliable REST and gRPC services, measurable model quality, and production-ready AI workflows.
