ML Engineering Intern — Oculi AI (Remote)
Worked on SPU-based computer vision pipelines for edge AI fall detection and conveyor belt anomaly monitoring, validating real-time detection models for production operation on resource-limited devices. Focused on hardware-software co-design to optimize model inference under constrained edge conditions. Contributed to deployment-aligned testing and evaluation for continuous production readiness. • Real-time validation of detection models on limited edge hardware • Edge AI pipeline development using Magenta simulator • Hardware-software co-design for privacy-preserving distributed AI deployment • Production readiness checks for operational accuracy and latency