1 day ago
Hanoi, Vietnam or Ho Chi Minh City, VietnamStaff+
Responsibilities
- Train and fine-tune large-scale LLMs and VLMs for multimodal and agentic tasks.
- Build end-to-end agentic AI prototypes for on-device and cloud deployment.
- Improve model efficiency through quantization, pruning, and distillation.
- Collaborate with hardware, software, and systems teams to bring research into products.
- Implement experiments in Python and PyTorch and manage benchmarks across large-scale distributed setups.
- Share knowledge, mentor junior engineers, and support cross-functional research initiatives.
Requirements
- Bachelor's, master's, or PhD in Computer Science, Machine Learning, Engineering, Information Systems, or a related field.
- At least 3 years of hands-on experience designing, training, or optimizing LLMs, VLMs, or foundation models in real-world settings.
- Minimum qualification paths require 4+ years with a bachelor's degree, 3+ years with a master's degree, or 2+ years with a PhD in hardware, software, systems engineering, or related work.
- Strong understanding of generative AI, multimodal architectures, and large-scale training strategies.
- Proficiency in Python and PyTorch, with experience using distributed training frameworks such as DeepSpeed or FSDP.
- Experience applying quantization, pruning, or distillation to reduce model footprint without degrading quality.
- Familiarity with hardware-aware AI optimization and deploying models on edge devices is preferred.
- Publications at leading ML/AI venues such as NeurIPS, ICML, CVPR, or ACL are preferred.
- Experience mentoring engineers or leading cross-functional research initiatives is preferred.
Categories
AI ResearchML Engineering
About Qualcomm
Qualcomm is a public semiconductor company headquartered in San Diego, founded in 1985, that designs and sells wireless chipsets and platforms for mobile devices, automotive, IoT, and networking, notably the Snapdragon application processors and 5G modems. It also licenses a large portfolio of cellular patents to device makers, generating revenue alongside chip sales; its technology underpins many Android smartphones and emerging automotive and edge-compute systems, and it trades on NASDAQ as QCOM.