Staff AI Engineer (Web3)

TL;DR

Staff AI Engineer (Web3): Designing and optimizing post-training pipelines for large language models with an accent on alignment, controllability, and reasoning capabilities. Focus on implementing DPO, GRPO, and RLAIF systems to improve model performance and production-grade inference deployment.

Location: APAC

Company

OKX is a leading crypto exchange and developer of OKX Wallet, providing global access to crypto trading and decentralized applications.

What you will do

• Lead the full post-training pipeline for LLMs, including supervised fine-tuning and preference optimization.
• Implement advanced training paradigms such as DPO (Direct Preference Optimization) and GRPO (Generalized Reward Policy Optimization).
• Develop domain-specific data recipes, curation strategies, and augmentation pipelines.
• Build and refine Reward Models and RLAIF closed-loop systems.
• Optimize inference efficiency and deploy models using low-latency frameworks like vLLM and SGLang.
• Evaluate model performance through automated benchmarks and human/AI feedback loops.

Requirements

• Bachelor's degree in Computer Science, AI, Machine Learning, or a related field.
• At least 8 years of industry experience.
• Hands-on experience across the full post-training pipeline for large models.
• Deep familiarity with preference learning and alignment techniques (DPO, GRPO, RL-based methods).
• Proven experience training specialized small models from scratch.
• Experience deploying models in production using vLLM, SGLang, or similar frameworks.

Culture & Benefits

• Competitive total compensation package.
• L&D programs and education subsidies for professional growth.
• Comprehensive healthcare schemes for employees and their dependants.
• Wellness and meal allowances.
• Various team building programs and company events.