Senior Agentic AI Engineer – Remote

Overseas; La Pampa; Remote; ARS 182,437,000 - 364,875,000; Full time

Job summary

Tether is seeking an AI Research Engineer (Agentic Posttraining) to advance post‑training methodologies, focusing on agentic behaviors and tool use across edge devices and cloud environments.

You will design multi‑modal post‑training systems, curate training data, and enhance reasoning and tool invocation. Collaboration with cross‑functional teams is essential to push SOTA results in a fully remote, globally distributed setup.

Qualifications

• Degree in CS or ML, or related field; MS/PhD preferred.
• Experience with multimodal post-training workflows.
• Experience applying post-training at scale using distributed training frameworks.
• Strong English communication skills.

Responsibilities

• Conduct end-to-end research and engineering to advance post-training of agentic and tool-use models to achieve SOTA results.
• Drive cross-cutting model improvements: factuality, instruction adherence, tool use, multi-agent coordination, reasoning calibration.
• Design and enhance large-scale post-training systems: data pipelines, training workflows, evaluation frameworks, benchmark infrastructure.
• Develop robust evaluation suites and diagnostic tools for deployment readiness.
• Strengthen feedback loops from real-world product usage and user signals into post-training.
• Collaborate with tooling, product, and training teams to improve agentic capabilities of frontier models.
• Determine production-ready integrations for major model releases.

Skills

Multimodal ML

Agentic post-training

Distributed training

Tool use integration

Education

Degree in CS/ML or related field

Job description

Tether is seeking an AI Research Engineer (Agentic Posttraining) to advance post‑training methodologies, focusing on agentic behaviors and tool use across edge devices and cloud environments.

You will design multi‑modal post‑training systems, curate training data, and enhance reasoning and tool invocation. Collaboration with cross‑functional teams is essential to push SOTA results in a fully remote, globally distributed setup.