AI Research Engineer (Agentic Post-training)

New

TetherArtificial intelligence

100% remote work – work from anywhere in the worldFull-Time

Salary not disclosed

Apply NowOpens the employer's application page

Job Details

Have a strong background in machine learning and hands-on experience training and fine-tuning large language models.
Have a deep understanding of post-training methods, including SFT, RLHF, DPO, and reinforcement learning.
Have experience building evaluation frameworks and benchmarks for LLM capabilities.
Be proficient in Python and modern ML frameworks such as PyTorch, JAX, or similar.
Have experience with distributed training and large-scale experimentation.
Be familiar with agentic systems, tool use, and multi-agent architectures.
Be able to work independently in a remote, asynchronous environment.
Have working proficiency in English.

Design, implement, and evaluate post-training recipes for large language models, including SFT, RLHF, DPO, and agentic reinforcement learning.
Build evaluation harnesses and benchmarks for agentic capabilities, including tool use, planning, memory, and multi-step reasoning.
Develop data pipelines to collect, clean, and curate training and evaluation data from real-world interactions.
Run large-scale experiments and analyze results, iterating on model behavior, prompts, and training configurations.
Collaborate with product and engineering teams to integrate research prototypes into production systems.
Stay current with research in LLMs, agentic AI, and reinforcement learning, and share findings with the team.
Contribute to internal tools and infrastructure for model training, evaluation, and deployment.
Publish research findings and represent Tether Data at conferences and in the open-source community, where appropriate.

View Full Description & ApplyYou'll be redirected to the employer's site