Appen Unveils New Platform for Customizing Large Language Models

Appen Limited (ASX: APX), a global leader in data solutions for AI development, has announced the launch of its latest platform capabilities designed to assist enterprises in customizing large language models (LLMs). This new solution aims to streamline the process for businesses looking to integrate generative AI into their operations, facilitating the development of AI applications with greater efficiency and accuracy.

Empowering enterprises with tailored AI solutions

Appen’s new platform capabilities offer internal teams a comprehensive framework for customizing LLMs according to their specific requirements. Through a seamless process embedded within Appen’s AI Data Platform, users can navigate the entire lifecycle of LLM model development, from initial selection to production deployment.

The platform provides a series of essential steps to guide users through the customization process:

Users can connect to various existing models, evaluate their performance, and conduct thorough benchmarking to identify the most suitable option for their needs.

High-quality data is crucial for training accurate AI models. Appen’s annotation platform enables the preparation of datasets essential for vectorization and Retrieval-Augmented Generation (RAG). Custom prompts are vital for validating model performance across different use cases. Appen’s platform facilitates the creation of tailored prompts either by leveraging internal expertise or tapping into a global crowd of contributors.

The platform streamlines the collection of human feedback for evaluating model performance, offering templates for human evaluation, A/B testing, and benchmarking to ensure optimal results throughout the development process.

Appen prioritizes model safety, providing detailed workflows and dedicated teams to identify and mitigate potential risks such as toxicity, brand safety issues, and other forms of harm.

Enabling enterprise innovation

Ryan Kolln, CEO of Appen, emphasized the significance of generative AI in driving enterprise innovation while acknowledging the challenges associated with ensuring the accuracy and trustworthiness of LLM-enabled applications. With over 25 years of experience in human-AI collaboration, Appen aims to empower enterprises to build and deploy reliable LLM applications efficiently.

Appen’s new capabilities enable enterprises to leverage both curated crowd-sourced data and proprietary datasets, ensuring optimal LLM output tailored to their specific use cases. This flexibility allows companies to strike a balance between model accuracy, complexity, and cost-effectiveness, whether deploying solutions on-premises, in the cloud, or through hybrid environments.

About Appen

Appen (ASX: APX) is a global leader in data solutions for the AI lifecycle, boasting over 25 years of experience in data sourcing, annotation, and model evaluation. With a global crowd of over 1 million contributors speaking more than 235 languages, Appen enables organizations to launch innovative AI products with speed and scale across various industries, including technology, automotive, finance, retail, healthcare, and government.

Appen’s latest platform capabilities represent a significant advancement in the field of AI customization, offering enterprises a streamlined approach to developing and deploying LLM-enabled applications with accuracy and trustworthiness. With its extensive expertise and robust infrastructure, Appen continues to drive innovation and empower businesses worldwide in their AI endeavors.