All Products
Search
Document Center

Platform For AI:EAS use cases

Last Updated:Mar 10, 2025

This topic lists the use cases of Elastic Algorithm Service (EAS) of Platform for AI (PAI).

LLM

Case

Description

Quickly deploy LLMs in EAS

Use EAS to deploy a large language model (LLM) application with WebUI and API support.

AIGC

Case

Description

Quickly deploy an MLLM in EAS

Deploy and call a MLLM inference service using EAS.

Use ComfyUI to deploy an AI video generation model service

Deploy an AI video generation application with ComfyUI and Stable Video Diffusion in EAS. A quick text-to-video solutions for creating short videos or animations on social platforms.

Use Stable Diffusion web UI to deploy an AI painting service

Deploy Stable Diffusion as a web application in EAS and generate images from text prompts.

Deploy a LoRA SD model by using Kohya_ss in EAS

Deploy open-source Kohya_ss using EAS and train LoRA Stable Diffusion model.

RAG

Case

Description

RAG-based LLM chatbot

The complete development process for a RAG-based LLM chatbot.

Use EAS and Elasticsearch to deploy a RAG-based LLM chatbot

Integrate Elasticsearch with RAG services. It also includes the core capabilities of the RAG chatbot and the unique features of Elasticsearch.

Use EAS and OpenSearch to deploy a RAG-based LLM chatbot

Integrate OpenSearch Vector Search Edition with RAG services. It also includes the core capabilities of the RAG chatbot and the unique features of OpenSearch Vector Search Edition.

Use EAS and ApsaraDB RDS for PostgreSQL to deploy a RAG-based LLM chatbot

Integrate an ApsaraDB RDS for PostgreSQL instance with RAG services. It also includes the core capabilities of the RAG chatbot and the unique features of ApsaraDB RDS for PostgreSQL.

EAS general solutions

Case

Description

Best practices for accessing dedicated gateway across VPCs

Enable cross-VPC access through a dedicated gateway using Cloud Enterprise Network (CEN).

Deploy a Hugging Face model in EAS

Deploy a Hugging Face model as a service application in EAS.

Best practices for PAI-EAS Spot

Best practices for using PAI-EAS Spot to optimize resource utilization, reduce costs, and maintain service stability.

Enable the tracing analysis feature in EAS

Implement link tracing features when deploying services in EAS.