Member of Technical Staff (Model Quantization & Inference)
A role with this title usually asks for solid Python and deep learning framework skills, plus hands on work shrinking models through quantization and getting them to run fast at inference. A strong candidate brings care with numerical accuracy, benchmarking discipline, and the ability to explain tradeoffs between speed, size and quality. At a company hardcoding model weights into silicon, that work has to line up closely with hardware constraints.
How it fits the AI era
This is core AI infrastructure work, so the role sits close to where models meet hardware. AI tools speed up experimentation and benchmarking, but judging whether a quantized model still behaves well and spotting subtle numerical failures stays human. The craft of making inference fast and predictable is likely to matter more, not less, as models spread.
How it builds AI experience
The role builds direct experience in quantization methods, inference optimization and hardware aware model design, which are in demand across the AI stack. A candidate can strengthen this by reproducing quantization results on open models, measuring accuracy and latency, and studying how fixed point and low precision formats behave. Working near custom silicon adds a rare view of how model choices affect real hardware.
Worth showing
- Python and PyTorch
- Inference benchmarking
- Numerical accuracy checks
- Hardware aware optimization
This is our own read, written from the employer’s posting and the company’s Spotlight. The employer’s posting is the source of truth for the role.
About Lamb Labs
Lamb Labs develops custom model processing units that hardcode artificial intelligence model weights into silicon for fast inference, per the company.
#J-18808-Ljbffr…