ONSITE or HYBRID at Albs
Core
Building real-time multimodal models (text, speech, vision) that run on-device with optimized architecture, inference, and runtime.
Role type
Senior/Staff Member of Technical Staff (Research or Engineering)
Builds
On-device multimodal AI models, inference engines, and agentic platforms
Domain
AI research and on-device machine learning
Deliverable
production ML models
Required skills
LLM pre-training and post-training, speech and vision-language modeling, model adaptation, inference optimization, hardware optimization (C++/Rust, CUDA, SIMD, NPUs), data pipeline construction
Preferred skills
Experience with agentic platforms, large-scale training on dedicated compute
Technologies
C++, Rust, CUDA, SIMD, NPUs
Responsibilities
Designing model architectures for latency and power efficiency, optimizing inference runtimes, building data pipelines for multimodal data, publishing research at top venues
Seniority
Senior/Staff, hands-on IC