Applied Research Engineer (AI)

TL;DR

Applied Research Engineer (AI): Developing advanced systems and methods for human-in-the-loop data and model alignment with an accent on RLHF, DPO, and data quality measurement. Focus on designing AI-assisted labeling tools, bridging research breakthroughs with real-world applications, and scaling data strategies for frontier models.

Location: Hybrid model with 2 days per week in office in San Francisco (US) or Wrocław (Poland)

Salary: $250,000–$300,000 USD

Company

hirify.global provides integrated infrastructure for frontier AI development, including annotation tools, data labeling services, and an expert marketplace for training data.

What you will do

• Develop and implement advanced AI alignment methods such as RLHF and DPO.
• Design systems to measure and improve human-in-the-loop data quality.
• Create AI-assisted tools leveraging active learning to enhance labeling efficiency.
• Investigate how various feedback types impact model performance.
• Integrate research breakthroughs into the product suite for scalable real-world application.
• Engage with the AI community through publications and thought leadership.

Requirements

• Ph.D. or Master’s degree in Computer Science, Machine Learning, or AI.
• 3+ years of experience in solving complex ML challenges.
• Strong foundation in frontier AI models and human data strategies.
• Proficiency in Python and deep learning frameworks like PyTorch, JAX, or TensorFlow.
• Proven track record of publishing in top-tier AI/ML conferences.
• Must be able to work in a hybrid model in San Francisco or Wrocław.

Nice to have

• Experience in human-computer interaction (HCI) research.
• Background in AI ethics and safety research.

Culture & Benefits

• Fast-paced startup-like environment focused on high agency and impact.
• Collaborative environment with opportunities for continuous learning.
• Career growth directly tied to individual contributions and ownership.
• Engagement in cutting-edge research with the opportunity to publish work.