Member of Technical Staff, Mid-training (AI)

Location: San Jose, United States

Salary: $180,000–$450,000 annually base salary

Company

hirify.global is an artificial intelligence company developing personalized, multimodal intelligence and next-generation hardware for natural interaction between people and machines.

What you will do

• Design and implement mid-training strategies for reasoning, planning, tool use, and long-horizon decision-making.
• Scale synthetic data generation pipelines for coding, agent trajectories, and multimodal data, and optimize data mixtures for reinforcement learning.
• Build and optimize distributed training pipelines for large models across GPU clusters.
• Develop evaluation frameworks for task success, reasoning quality, and tool-use accuracy.
• Run experiments and ablations to analyze training dynamics, scaling behavior, and bottlenecks.
• Collaborate with pre-training, post-training, and product teams on model development and agent use cases.

Requirements

• Strong machine learning background with hands-on experience training or fine-tuning large language, multimodal, or equivalent models.
• Deep understanding of reinforcement learning, including policy optimization, reward design, exploration, and environment design.
• Experience with simulation or execution environments such as code interpreters, sandboxed execution, game environments, or robotics simulators.
• Ability to design rigorous experiments and diagnose training failures and scaling bottlenecks.
• Proficiency in Python and PyTorch, with comfort working across research and systems code.
• Ability to work in a fast-moving, research-focused environment with uncertain approaches.

Nice to have

• Experience with mid-training, post-training, agent-focused model development, or coding LLM training.
• Familiarity with synthetic data, trajectory-based training, or model distillation pipelines.
• Experience training or scaling models with 100B+ parameters or equivalent systems.
• Open-source ML contributions or publications at leading machine learning conferences.
• Experience optimizing distributed training systems, GPU utilization, memory efficiency, or communication.

Culture & Benefits

• Research-forward environment focused on developing new model capabilities.
• Cross-functional collaboration across pre-training, post-training, and product development.
• Full-time compensation may include additional components and benefits beyond the base salary.