AI Research Engineer (Agentic Post-training) - 100% Remote Worldwide

Tether Operations Limited

21h • Remote

About The Position

As a member of Tether’s AI model team, the AI Research Engineer will advance post-training methodologies, particularly for agentic behaviors and tool use. The role focuses on refining pretrained models to improve intelligence and domain-specific capabilities while enabling them to reason, plan, and autonomously invoke external tools to solve real-world, multistep tasks, including on edge devices such as smartphones.

The work spans resource-efficient agents for limited hardware and complex multimodal architectures integrating text, images, and audio. The engineer will develop and implement research-driven post-training algorithms, curate agentic training data, improve baseline performance, and resolve bottlenecks to achieve state-of-the-art quality in tool-augmented agents. The role also involves building post-training systems and evaluations, incorporating real-world product feedback, and collaborating with research, engineering, tooling, product, and training teams to prepare production-ready integrations for model releases.

Requirements

• Degree in Computer Science, Machine Learning, or a related field.
• Experience with multimodal post-training workflows and data pipelines, particularly for agentic systems and tool use.
• Hands-on experience applying post-training at scale using distributed training frameworks, such as multi-node GPU environments.
• Demonstrated experience improving model capabilities in areas such as reasoning, tool use, and multi-agent coordination to achieve state-of-the-art results.
• Proven open-source contributions related to agentic systems or tool use, such as code, datasets, or models on platforms including GitHub or Hugging Face.

Nice To Haves

• Advanced degree (MS/PhD).
• Strong publication record at top-tier AI conferences.
• Publications at leading AI conferences such as NeurIPS, ICML, ICLR, ACL, CVPR, or ECCV.

Responsibilities

• Conduct end-to-end research and engineering initiatives to advance post-training of agentic and tool-use models and achieve state-of-the-art results.
• Drive model improvements in factuality, instruction adherence, tool and function use, multi-agent coordination, and reasoning calibration.
• Design and enhance large-scale post-training systems, including data pipelines, training workflows, evaluation frameworks, and benchmark infrastructure.
• Develop rigorous evaluation suites and diagnostic tools to assess model readiness for deployment.
• Strengthen feedback loops from real-world product usage by incorporating explicit and implicit user signals into post-training.
• Collaborate with tooling, product, and training teams to improve the usefulness, reliability, and agentic capabilities of frontier models.
• Liaise with research, engineering, and cross-functional teams to determine which integrations are production-ready for major model releases.
• Curate agentic training data, including tool-use trajectories, reasoning chains, and environment interactions.

What This Job Offers

• Job Type: Full-time
• Career Level: Mid Level