TELUS International Launches Fine-Tune Studio to Deliver High-Quality Datasets that Improve the Performance, Adaptability and Safety of Generative AI Models

Multimodal, multi-language workflows optimize AI dataset creation processes to help clients get accurate fine-tuning data, predictably fast, at affordable costs

TELUS International, a leading digital customer experience (CX) innovator that designs, builds, and delivers next-generation solutions, including artificial intelligence (AI) and content moderation, today announced the launch of Fine-Tune Studio, a task-execution platform designed to create high-quality fine-tuning datasets for large language models (LLMs) and generative AI (GenAI) models. Fine-Tune Studio will be presented during the company’s attendance at the 2024 World Summit AI in Montreal, Canada on April 24-25.

Fine-Tune Studio is integrated in TELUS International’s fully-managed AI Data Platform as part of the company’s fully-managed GenAI data services available to LLM builders and adopters. Fine-Tune Studio creates high-quality datasets for LLM fine-tuning in more than 100 languages and can be deployed across domains where specialized responses are required, including science, technology, engineering and mathematics (STEM), healthcare, finance, law, coding and more. Aligning with the paradigm shift to multimodal AI use cases, Fine-Tune Studio supports labeling tasks combining various data types like text, audio, image, and video for supervised fine-tuning (SFT), reinforcement learning through human feedback (RLHF), direct preference optimization (DPO) and red teaming.

“As AI continues to be incorporated into the fabric of society and our daily lives, it is more critical than ever that companies make responsible decisions to ensure the accuracy, diversity and integrity of the datasets used to train their models when scaling their GenAI use cases from concepts to pilots to production and beyond,” said Siddharth Mall, Vice President, Product for TELUS International AI Data Solutions. “With the launch of Fine-Tune Studio, we have taken another significant step forward in our ongoing efforts to develop proprietary technology that supports our ability to deliver expertly-vetted and validated data that improves our clients’ model performance, adaptability, and safety.”

Fine-Tune Studio’s native AI integrations include real-time plagiarism, tone analysis, spelling and grammar checks to improve output accuracy and facilitate users’ task completion, driving increased productivity and efficiency. Configurable and customized task and quality assurance workflows provide easy project management for clients to find the most suitable quality control processes, such as consensus, algorithmic and deterministic, human-in-the-loop, and probabilistic methods.

Marketing Technology News: Innovid’s Annual CTV Insights Report Finds Interactive CTV Campaigns Drive Up to 10.3x Higher Engagement for Advertisers

Experts Engine: the right people working on the right tasks at the right time

The growing need for industry-specific and highly-specialized training datasets for increasingly complex AI use cases has elevated the importance of using subject matter experts to create, label and validate data. TELUS International’s Fine-Tune Studio task-execution platform works in conjunction with Experts Engine, another proprietary feature of its AI Data Platform that algorithmically matches tasks with best-fit experts from the company’s global crowdsourced AI community of more than one million contributors.

In addition to generalists, TELUS International’s AI Community includes linguists, researchers, coders and doctors as well as experts from various specialty domains. These individuals are able to capture more complex layers of information and recognize subtle but crucial distinctions while working on data tasks like response generation, response ratings and model interaction. The use of experts mitigates the creation of inaccurate and incomplete datasets that can mislead the AI during training, causing it to learn incorrect patterns, make faulty assumptions and introduce unintentional biases.

“With a keen focus on data integrity, we empower our clients to forge a future in which technology not only augments human potential but does so with unwavering precision and a steadfast commitment to ethical and regulatory standards,” concluded Mall.