Job Description
Job Description
Position: Applied Research Engineer
Location - San Francisco, CA - Hybrid
Job Type: Full Time
Job Description:
Role Overview
As an Applied Research Engineer at Labelbox, you’ll play a critical role in shaping the future of human-in-the-loop AI systems. You’ll design and implement advanced methods to align human feedback with the training of cutting-edge AI models, including techniques like Reinforcement Learning from Human Feedback (RLHF), Direct Preference Optimization (DPO), and other alignment strategies. You’ll also develop innovative tools to measure and enhance the quality of human-generated data and build AI-assisted systems that streamline and improve the data labeling process.
Your work will directly influence the performance and reliability of frontier models by ensuring they reflect human preferences more accurately. By bridging research with real-world applications, you’ll help bring scalable, impactful alignment solutions into production for some of the world’s most advanced AI developers.
Your Impact
- Drive advances in AI alignment, developing state-of-the-art techniques like RLHF and new methods to ensure models align with human intent.
- Improve human-in-the-loop data quality by building robust systems for measurement, feedback analysis, and refinement.
- Create AI-assisted labeling tools using active learning, adaptive sampling, and automation to enhance speed and accuracy.
- Investigate the role of different feedback types—like preferences, critiques, and demonstrations—on model training and alignment.
- Develop algorithms to improve how AI systems learn from human input, increasing adaptability and response quality.
- Integrate alignment innovations into the Labelbox product suite, making human feedback workflows scalable for customers.
- Collaborate with customers and the AI research community to shape best practices for training large-scale models.
- Contribute to the field through publications, conference presentations, and open research.
- Stay ahead of the curve by exploring emerging trends in AI alignment, data quality, and human-AI collaboration.
- Help establish Labelbox as a thought leader in human-centric AI by producing high-quality technical and educational content.
What You Bring
- Advanced degree (Ph.D. or Master’s) in Computer Science, Machine Learning, AI, or a related field.
- 3+ years of experience solving complex ML challenges and building production-ready AI systems.
- Proven track record in data quality systems, with a strong understanding of how they affect model performance.
- Deep knowledge of frontier models, including large language models and multimodal systems, and how to optimize them using human feedback.
- Strong programming skills in Python, with hands-on experience in PyTorch, JAX, or TensorFlow.
- Contributions to the research community through peer-reviewed publications in top AI/ML conferences (e.g., NeurIPS, ICML, ICLR, ACL, EMNLP, NAACL).
- Ability to translate research into real-world solutions, from prototype to scalable deployment.
- Strong analytical thinking and a structured approach to solving open-ended AI problems.
- Excellent collaboration and communication skills to work effectively across technical teams and with external stakeholders.