Agentic AI Engineer — Post-Training & Tooling (Remote)

Puente Talent Partners; Tandil; Remote; ARS 185,448,000 - 278,173,000; Full time

Job summary

Tether is seeking an AI Research Engineer (Agentic Posttraining) to advance post-training methods for agentic AI. You will refine pre-trained models to reason, plan, and autonomously invoke external tools for real-world, multi-step tasks on edge devices, including smartphones.

You will work with systems ranging from lightweight agents to complex multi-modal architectures, optimized for tool-augmented decision making. Strong expertise in LLMs and agentic workflows is required.

Qualifications

• Deep expertise in large language model architectures.
• Experience with post-training for agentic workflows.
• Proficiency in tool-use fine-tuning and function calling.
• Familiarity with reinforcement learning from feedback on multi-turn interactions.
• Ability to curate agentic training data and improve model performance.

Responsibilities

• Drive innovation in post-training methodologies for agentic AI.
• Develop and evaluate tool-augmented agents and multi-modal systems.
• Curate trajectories, reasoning chains, and environment interactions for training data.

Skills

LLM architectures

Agentic post-training

Tool use fine-tuning

Reinforcement learning from feedback

Edge-device inference

Tools

Large-scale ML frameworks

Data curation pipelines

Job description

Tether is seeking an AI Research Engineer (Agentic Posttraining) to advance post-training methods for agentic AI. You will refine pre-trained models to reason, plan, and autonomously invoke external tools for real-world, multi-step tasks on edge devices, including smartphones.

You will work with systems ranging from lightweight agents to complex multi-modal architectures, optimized for tool-augmented decision making. Strong expertise in LLMs and agentic workflows is required.