Deep Learning Research Intern
Skip the busywork
ApplyBolt rewrites your resume for this exact role and hits submit. You just pick the jobs.
About this role
(Embodied AI, Multimodal Foundation Models & Efficient Systems)
About Us
Futurewei is a well-funded independent research organization with a long history of R&D innovation in Silicon Valley. We are committed to open-source development, fundamental research, and advancing next-generation intelligent systems through collaboration and standards development.
About the Role
We are seeking a strong deep learning research intern to join our ASID team in San Jose, CA. This role focuses on building learning systems for embodied intelligence, emphasizing how multimodal foundation models can be trained, compressed, and deployed efficiently in embodied and interactive environments.
Our work goes beyond static perception. We study intelligence grounded in embodied experience—the interaction of perception, action, and environment over time—while ensuring models remain efficient, scalable, and deployable in real-world systems.
Core Research Focus Areas
The intern will contribute to one or more of the following interconnected research directions:
1. Multimodal Foundation Models
Fine-tuning and adaptation of large language models (LLMs), vision-language models (VLMs), and vision-language-action (VLA) models
Multimodal representation learning across vision, language, and action
Grounding foundation models in embodied experience and temporal interaction
2. Neural (Generative) Image and Video Compression
Learning-based image and video compression models
Efficient visual representations for perception and downstream embodied tasks
Joint optimization of compression efficiency, reconstruction quality, and task relevance
3. Embodied AI
Learning frameworks that couple perception, action, and environment dynamics
World models, predictive learning, and agent-centric representations
Embodied learning in simulation or real-world-inspired environments
4. Model Compression & Inference Acceleration for Embodied Systems
Model compression, pruning, quantization, and distillation
Efficient inference and deployment strategies for embodied and real-time applications
Hardware- and system-aware optimization for edge or robotic platforms
Responsibilities
Conduct research in one or more of the focus areas above
Design and implement learning algorithms and experimental pipelines
Develop prototype systems or demos for embodied and multimodal AI applications
Collaborate closely with researchers in a fast-paced, research-driven environment
Qualifications
MS or PhD in Computer Science, Electrical Engineering, Artificial Intelligence, Robotics, Mathematics, or a related field
Strong foundation in machine learning and deep learning
Experience or strong interest in multimodal models, embodied AI, compression, or efficient inference
Proficiency with PyTorch; experience with HuggingFace or similar frameworks is a plus
Solid Python programming skills
Research experience with publications in top conferences or journals preferred
Strong communication skills and ability to work effectively in a global research team
Location: San Jose, CA
Hourly interns pay range: $18 to $59, depending on degree-seeking academic program (PhD, Master’s, Bachelor’s, etc.), years of relevant experience, year in school, geographic location, credentials, qualifications, and other job-related factors.
Housing allowance and relocation benefit might be provided to intern candidates who meet the qualifications. Additional details on the compensation package will be provided to candidates during the interview process.