Reinforcement Learning Engineer (AI)

TL;DR

Reinforcement Learning Engineer (AI): Developing and optimizing advanced RL workflows for self-improving agents with an accent on data efficiency, training methodologies, and large-scale model optimization. Focus on solving complex research challenges in sequence-level importance ratios and continuous learning, leveraging massive GPU infrastructure to push the boundaries of AI agents.

Location: Remote within the US or hybrid (San Francisco/Bellevue). Must be a U.S. person or eligible for export-controlled information access.

Salary: $188,000–$275,000

Company

hirify.global is an AI Hyperscaler providing high-performance GPU infrastructure and end-to-end platforms for developers to build, train, and scale AI models.

What you will do

• Generate and investigate research ideas to overcome obstacles in continuous learning.
• Collaborate with the OpenPipe team to validate research directions on real-world customer tasks.
• Engineer scalable and data-efficient training methods for self-improving agents.
• Utilize massive compute clusters to run large-scale training experiments and ablations.
• Build and maintain production services supporting the RL platform lifecycle.

Requirements

• Proven experience training LLMs to SOTA on specific tasks.
• Strong engineering background with a focus on high-impact project results.
• Deep understanding of RL methods, specifically sequence-level or token-level importance ratios.
• Eligibility for access to export-controlled information (U.S. person or authorized status required).
• Proficiency in Python and familiarity with the stack (Kubernetes, Postgres, FastAPI).

Culture & Benefits

• 100% employer-paid medical, dental, and vision insurance.
• 401(k) retirement plan with a generous employer match.
• Comprehensive disability, life insurance, and mental wellness support (Spring Health).
• Flexible Paid Time Off policy to support work-life balance.
• Paid parental leave and family-forming support through Carrot.
• Hybrid work environment with quarterly team gatherings for collaboration.