Agentic AI Researcher
Core
Designing and building next-generation agentic systems that enable personalized user experiences by leveraging visual reasoning and multimodal models.
Role type
Senior IC agentic AI researcher (multimodal systems)
Builds
Production-grade agentic systems operating over user visual data, hybrid routing systems for on-device and cloud VLMs, and agent tools for generative models.
Domain
Artificial Intelligence, Multimodal Machine Learning, Computer Vision, Agentic Systems
Deliverable
production ML models
Required skills
Python, PyTorch, agentic multimodal LLM systems, multi-agent system orchestration, reinforcement learning for agents, visual perception, VLMs, hybrid routing design
Preferred skills
Applied research in industry, publications in NeurIPS/CVPR/ICCV/ECCV, product-focused applicative research, hands-on experience building systems powered by Vision and LLMs
Technologies
Python, PyTorch, VLMs, multimodal models
Responsibilities
Design and prototype agentic architectures for visual data; track and adapt latest research into production systems; explore visual perception approaches for video understanding and spatio-temporal grounding; investigate tradeoffs between on-device and cloud models for hybrid routing; collaborate on underlying generative models; partner with product/design to translate creative requests into agent behaviors.
