CareerPlan

Member of Technical Staff, Pretraining Architecture and Scaling

HQFull-time$180k–$450k2026-11-06 → 2026-11-19

Core

Building multimodal foundation models (text, audio, vision) and pretraining infrastructure for agentic AI systems.

Role type

Senior IC machine-learning engineer (multimodal pretraining)

Builds

Multimodal foundation models and training infrastructure

Domain

Artificial Intelligence / Machine Learning

Deliverable

production ML models

Required skills

Large model training, scaling laws, distributed training, model architecture research, experimental design, debugging training workloads

Preferred skills

Mixture-of-experts, sparse architectures, long-context training, kernel-level optimization, mixed-precision training

Technologies

GPU clusters, distributed training frameworks

Responsibilities

Research model architecture and optimization; Run controlled experiments to evaluate scaling; Develop training recipes (learning rates, context lengths); Investigate MoE and hybrid attention; Diagnose training instability and numerical issues

Seniority

Senior, hands-on IC