Experienced with LLM training methodologies? Eager to put that knowledge to work with leading LLM foundational model builders? Join TELUS International AI Data Solutions and help bring the latest generative AI technologies to the next level of performance and accuracy! We’re hiring experienced supervised fine-tuning and reinforcement practitioners to work with A-list clients to help shape the future of AI.

Job Description

This is a customer-facing subject matter expert role, consulting with and advising our clients and prospects on optimal Generative AI supervised fine-tuning and reinforcement learning workflows and training datasets. You’ll review and advise on annotation guidelines, production workflows, tool specifications, quality assurance methods, and performance metrics. The position is cross-functional, advising our operations and product teams on best practices for pilots and at-scale programs.

Responsibilities

• Work with clients and prospects to understand AI training dataset needs

• Design task execution workflows for building SFT and RLHF datasets, based on customer and prospect requirements

• Develop red teaming methodologies aligned to customer and prospect content policies and business objectives

• Specify production workforce profiles that meet project requirements

• Establish quality control methodologies that meet quality metric targets

• Train project managers in SFT, RLHF and red teaming best practices

• Analyzes reporting needs and proactively creates new insights

• Support continuous process improvement for operations, workforce sourcing and quality control methodologies

Qualifications / Requirements

• BA or BSc in Computer Science, Software Engineering, Machine Learning or related field

• 5+ years work experience in AI fields, with 1+ years work experience in Generative AI training technologies

• Strong client and team communication skills

• Exceptional analytical skills

• Strong organizational skills including prioritizing, time management, and ability to meet tight deadlines