TELUS International, an IT service company, has introduced “Fine-Tune Studio,” designed to create accurate and relevant datasets for large language models (LLMs) and GenAI models.

The platform, integrated into TELUS International’s AI Data Platform, offers services like supervised fine-tuning, reinforcement learning, and direct preference optimization across various domains. It manages labeling tasks across various data types such as text, audio, image, and video. Key features of this platform include real-time plagiarism checks, tone analysis, spellcheck and grammar corrections, and a workflow for task and quality assurance.

TELUS International claims Fine-Tune Studio ensures improved model performance, adaptability, and safety. The product is said to provide the ability to effectively manage complex data tasks using a diverse and specialized workforce.

Analyst QuickTake: This launch by TELUS International comes in the wake of firms like Open AI and Alphabet having introduced and popularized mass dataset creation for LLMs. This key development aligns with the shift towards multimodal AI use cases by enabling the labeling of tasks across various data types, ranging from texts to videos.