Magic · San Francisco +1 · Remote · Sep 19
Optimize throughput and latency for ultra-long-context training and inference.
$275k–$550k
Software EngineeringM
Magic
HiringFrontier labBuilding safe AGI that automates research and code generation to accelerate progress.
magic.dev/ ↗CHECKED 6H AGO
Open roles (8)
8 open roles
Magic · San Francisco · Onsite · Sep 19
Design and operate infrastructure training long-context models at scale.
$275k–$550k
Research EngineersMagic · San Francisco · Onsite · Oct 6
Build safe AGI to accelerate progress on humanity's most important problems.
$100k–$550k
Magic · San Francisco · Onsite · Sep 19
Design reward signals and training environments for long-horizon RL iteration.
$275k–$550k
Research ScientistsMagic · San Francisco · Onsite · Sep 19
Train trillion-parameter models on large GPU clusters
$275k–$550k
Research EngineersMagic · San Francisco · Onsite · Sep 19
Design and scale inference serving for long-context frontier models.
$275k–$550k
Inference EngineeringMagic · San Francisco · Onsite · Sep 19
Build security tooling protecting models, systems and engineering org.
$225k–$550k
Security EngineeringMagic · San Francisco · Onsite · Sep 19
Own the IT systems that keep the company productive and secure end-to-end.
$200k–$350k
Admin and ITLab profile
About the lab
Known for
- Frontier-scale pre-training
- Domain-specific RL
- Ultra-long context
- Inference-time compute
- Automating research and code generation