jobs
Inference Engineer, AGI
We are looking for an Inference Engineer to help serve inference for real-time multimodal conversational AI. This is a full-stack inference role: you will work across the path a model takes from…
In inglese
- Company
- Amazon
- Location
- Sunnyvale, California, USA
- Status
- Closed
- Posted
- 2026-10-05
Amazon is hiring an Inference Engineer in Sunnyvale for its AGI work on real-time multimodal conversational AI. The role spans model and inference co-design, low-latency streaming serving, GPU kernel and runtime work, and offline training/RL/evaluation systems. Responsibilities include quantization, speculative decoding, KV-cache management, continuous batching, multi-GPU inference, and benchmarking latency, throughput, and cost. The posting asks for a master’s or equivalent, at least three years of professional software development, and hands-on experience profiling and improving neural-model inference rather than only using frameworks.
Original job posting