Presentation
AI Research Scientist-Foundation Model Algorithms
SessionJob Postings
DescriptionWe are looking for a strategic technical leader to drive the development of the next generation of Huawei’s foundation models. This role focuses on algorithmic breakthroughs in large-scale models across NLP, multimodal, CV, and domain-specific verticals (e.g., document understanding, remote sensing, 3D).
You will lead research and engineering in model architecture, training methods, cross-modal alignment, in-context learning, hallucination mitigation, search augmentation, and agentic reasoning—driving capability from core (L0) infrastructure to application-level (L1) delivery.
Responsibilities:
1. Lead algorithm R&D for foundation models in NLP, vision, and multimodal domains (e.g., natural images, documents, 3D, OCR, remote sensing).
2. Drive innovation in key technical directions including encoder design, modality alignment, scalable training, prompt-based learning, hallucination control, search-augmented modeling, and agent-enablement.
3. Guide horizontal capability building at the L0 level and support model performance optimization and delivery in L1-level industry applications.
4. Interface with customers and internal stakeholders to gather requirements, define technical paths, and ensure model effectiveness aligns with business goals.
5. Contribute to roadmap planning, technical foresight, academic collaboration, and external technology scouting for NLP and multimodal models.
You will lead research and engineering in model architecture, training methods, cross-modal alignment, in-context learning, hallucination mitigation, search augmentation, and agentic reasoning—driving capability from core (L0) infrastructure to application-level (L1) delivery.
Responsibilities:
1. Lead algorithm R&D for foundation models in NLP, vision, and multimodal domains (e.g., natural images, documents, 3D, OCR, remote sensing).
2. Drive innovation in key technical directions including encoder design, modality alignment, scalable training, prompt-based learning, hallucination control, search-augmented modeling, and agent-enablement.
3. Guide horizontal capability building at the L0 level and support model performance optimization and delivery in L1-level industry applications.
4. Interface with customers and internal stakeholders to gather requirements, define technical paths, and ensure model effectiveness aligns with business goals.
5. Contribute to roadmap planning, technical foresight, academic collaboration, and external technology scouting for NLP and multimodal models.
Location
Shanghai, Shenzhen, Hong Kong, Singapore, Canada, Beijing, Hangzhou, etc.
In-Person:
Onsite
Description of Position:
PhD; Full-Time; Desired Level of Education: Master’s degree
Other Skills or Experience:
1. Deep understanding of the architecture, training, evaluation, and deployment of foundation models in NLP, CV, or multimodal AI.
2. Proven expertise in one or more verticals: classification, detection, segmentation, VLP, OCR, remote sensing, or 3D understanding.
3. Hands-on experience with model pretraining, fine-tuning, performance diagnosis, and architecture optimization (e.g., CNNs, Transformers).
4. Strong track record of driving core model R&D for mainstream products, with leadership in pre-research, architecture design, or technical delivery.
5. Demonstrated experience in scaling models from research to real-world applications, with deep insights into bottlenecks, evaluation, and business alignment.
6. Publications in top-tier conferences or journals are a strong plus (NeurIPS,SIGGRAPH,ICLR,ICML,CVPR etc).
·
Event Type
Job Posting
TimeSunday, 10 August 20258:00am - 9:00am PDT
LocationWest Building, Exhibit Hall C
Session TimeSunday, 10 August 20258:00am - 9:00am PDT
LocationWest Building, Exhibit Hall C
