Tagged: multimodal

Muhammad Umair is a Pakistan-based AI consultant, ML engineer, and PhD researcher who has spent seven-plus years building machine learning systems that actually ship. He leads AI training at atomcamp, has driven AI initiatives for UNDP Pakistan, and has built three AI SaaS products end-to-end. His PhD research at UESTC focuses on multimodal test-time adaptation and low-resource learning - the kind of work that makes AI usable in places where data is scarce and compute is expensive.