OKX is hiring a Staff AI Engineer, Model Post-Training and Alignment

OKX is hiring a hands-on ML engineer specialising in large-model post-training and alignment. The role covers the full post-training lifecycle — data strategy, reward modelling, RL-based optimization, and production inference — for the models behind one of the largest crypto exchanges and its wallet.

What you'll do

• Design, run, and optimize post-training pipelines that improve model performance, controllability, domain adaptation, and reasoning
• Own data strategy and reward modelling for alignment work
• Take models from RL-based optimization through to production-grade inference deployment

Requirements

• Deep, hands-on experience with large-model post-training: SFT, RLHF/RLAIF, or comparable alignment methods
• Strong applied ML engineering skills and comfort owning a pipeline end to end
• Track record of shipping models into production, not just research settings

Based in APAC. Full details and application at the link.