Remote LLM Agentic Trainer

Role Overview

We are looking for talented programmers with experience in any programming language or tech stack to join our team. Your primary responsibility will be to validate and ensure the quality of training data for developing agentic AI systems. This role involves utilizing function-calling methods to construct structured tasks and maintain data workflows that are consistent, accurate, and aligned with the latest AI requirements.

Key Responsibilities:

• Validate large-scale training datasets and uphold stringent quality standards.
• Create, manipulate, and structure JSON tasks for function-calling workflows.
• Collaborate closely with AI researchers and teams to ensure datasets meet agentic AI use case requirements.
• Monitor data accuracy, consistency, and integrity across various projects.
• Contribute to documentation and the development of best practices for dataset preparation and validation.

Requirements:

• Strong proficiency in Python, JavaScript, Java, or any other programming language suitable for data handling and workflow automation.
• Solid understanding of SQL for querying and managing datasets.
• Expertise or knowledge in handling JSON data.
• Exceptional attention to detail with a strong focus on data quality and consistency.
• Interest in or exposure to agentic AI, Large Language Models (LLMs), or Natural Language Processing (NLP) best practices.

Bonus Points:

• Prior experience with data annotation tools.
• Experience with prompt engineering for Agent AI.

Perks of Freelancing With Turing:

• Work in a fully remote environment.
• Opportunity to engage with cutting-edge AI projects and leading LLM companies.

Offer Details:

• Commitments Required: Minimum 4 hours per day and 20 hours per week, with a 4-hour overlap with Pacific Standard Time (PST). Options for 20, 30, or 40 hours per week are available.
• Engagement Type: Contractor assignment (no medical or paid leave benefits).
• Duration of Contract: 1 month, adjustable based on project needs.
• Location: Open to candidates in India, Pakistan, Nigeria, Kenya, Egypt, Ghana, Bangladesh, Turkey, Brazil, and Mexico.

Evaluation Process:

The evaluation involves an online assessment:

• Automated coding challenge (choose any programming language) - 30-45 minutes.
• Function calling MCQ test - 30 minutes.

Upon applying, you will receive an email with a login link to access the portal and complete your profile.