Alibaba Cloud, the digital technology and intelligence backbone of Alibaba Group, introduced a serverless version of its Platform for Enterprise Gen-AI (PAI)-Elastic Algorithm Service (EAS) during its AI & Big Data Summit in Singapore. This new service aims to provide a more cost-effective solution for model deployment and inference to both individuals and enterprises. Additionally, Alibaba Cloud integrated its vector engine technology into various product offerings, including Hologres, Elasticsearch, and OpenSearch, making it easier for enterprises to access large language models (LLMs) and build custom generative AI applications.
The PAI-EAS platform allows users to access computing resources as needed, eliminating the requirement to manage physical or virtual servers. Users will only be charged for the computing resources they use, resulting in a 50% reduction in inference costs compared to traditional pricing models. Currently undergoing beta testing, the serverless offering is accessible for image generation model deployment and is set to expand its capabilities in March 2024 to support the deployment of open-source LLMs and models from Alibaba’s Enterprise Gen-AI model community, ModelScope.
Alibaba Cloud’s technology advancements, including LLMs and vector engine technology, enable a Retrieval-Augmented Generation (RAG) process, allowing enterprises to enhance LLMs with their knowledge bases for improved outcomes. This leads to increased accuracy, faster retrieval of relevant information, and more nuanced insights, enhancing efficiency and decision-making across various applications.
Alibaba Cloud also unveiled an upgrade to its big data service, MaxCompute MaxFrame, which is a distributed Python data processing framework designed to meet the growing demand for data preprocessing and analysis in AI-related computing tasks. Additionally, they introduced PAI-Artlab, a platform for model training and image generation, aimed at fostering creativity among designers. This platform enables designers to produce professional-grade designs quickly and offers tools for model training without the need for coding skills.
In a significant move, Alibaba Cloud enhanced its range of database solutions by integrating its vector engine technology, improving performance and capabilities. This integration benefits tasks like similarity comparisons and semantic analysis, particularly benefiting LLMs and advancing various AI functionalities.
Global customers, including companies like Haleon and rinna, utilize Alibaba Cloud’s technologies to support their digital transformation efforts. For instance, Haleon collaborated with Alibaba Cloud to introduce an Enterprise Gen-AI nutritionist for Chinese consumers, while rinna leveraged Alibaba Cloud’s LLMs to enhance the performance of its models.
About Alibaba Cloud: Enterprise Gen-AI
Alibaba Cloud continues to lead in Enterprise Gen-AI and cloud technology innovation, offering solutions that empower enterprises with intelligence-driven capabilities for enhanced efficiency and performance.