Research Engineer / Research Scientist, Vision
Core
Research engineer developing visual and spatial reasoning capabilities for Claude LLMs through experiments, tool development, and evaluation.
Role type
Senior IC research engineer (computer vision & multimodal LLMs)
Builds
State-of-the-art Claude models with enhanced vision and spatial reasoning
Domain
Artificial Intelligence / Computer Vision / Large Language Models
Deliverable
production ML models
Required skills
Computer vision, software engineering, large vision language model architecture, synthetic and real-world visual dataset creation, systematic prompting, finetuning, evaluation
Preferred skills
Large-scale pretraining, supervised learning (SL), reinforcement learning (RL), deep learning on images/video, complex agentic systems, high-performance ML systems (GPUs/TPUs/JAX/PyTorch), large-scale ETL
Technologies
JAX, PyTorch, GPUs, TPUs
Responsibilities
Run experiments to evaluate architectural variants, data strategies, and SL/RL techniques; Develop and test tools and agentic infrastructure for visual reasoning; Create evaluations and benchmarks for multimodal capabilities; Partner with product org to solve API customer challenges related to vision and spatial reasoning
Seniority
Senior, hands-on IC
