This topic introduces Dify on DMS, a powerful AI development platform integrated into Alibaba Cloud Data Management (DMS). This feature lets you build custom AI applications, such as AI chat for data, enterprise knowledge bases, and SQL code generators, on your databases securely and efficiently without writing complex code or migrating data.
You own the container environment for your Dify instance.
Background information
Dify: A popular open-source, visual development platform for large language model (LLM) applications. It provides a full set of tools to create, orchestrate, and operate AI applications. These tools include prompt engineering, context management, and a retrieval-augmented generation (RAG) engine. For more information about Dify, see the official Dify documentation.
DMS: An enterprise-grade data management and development platform from Alibaba Cloud. It provides secure and efficient database management services.
Dify on DMS: A deep integration of DMS and Dify. This integration embeds the Dify platform directly into DMS. This allows your databases, such as MySQL, PostgreSQL, and SQL Server, to act as a secure and seamless knowledge source for Dify applications. It eliminates the need for intricate data synchronization pipelines - enable your AI applications to natively interpret your business data.
Core advantages
In addition to the features of the community edition of Dify, Dify on DMS provides the following advantages:
Advantage | Description |
Data security | AI applications access your data directly within the DMS environment. Data is not moved out of the database or exposed to the Internet. This provides maximum security for your core data assets. |
Zero data migration | No extract, transform, and load (ETL) or data synchronization is required. Connect directly to data sources managed by DMS for real-time, native access to your business data. This reduces architectural complexity and maintenance costs. |
Low-code development | Using the visual interface of Dify, business personnel or developers can quickly build AI applications like assembling building blocks. No deep background in algorithms is required. |
Versatile scenarios | Easily build various practical AI applications, such as natural language-to-SQL converters, AI chatbots, internal knowledge base Q&A systems, and financial report analysis bots. These applications help drive business innovation. |
Cost-effective | Reuse your existing DMS instances and database resources. This lets you explore and implement generative AI applications at a very low cost. |
Scenarios
AI data Q&A bot
Scenario: A marketing operations specialist wants to know, "Who was the top salesperson in the Shanghai region last quarter?" They can obtain the answer without asking a data analyst to write SQL.
Implementation: You can create a Q&A bot based on order and user tables. The operations specialist asks questions in natural language, and the AI automatically queries the database and returns the result.
Natural language to SQL generation
Scenario: A junior developer or data analyst needs to write a complex SQL query. This process is time-consuming and prone to errors.
Implementation: You can create an SQL generation tool. When you input "Query the average spending amount of all active users in the last 30 days," the AI automatically generates the correct SQL code.
Private enterprise knowledge base
Scenario: You need to build an intelligent Q&A system for employees or customers that uses content stored in a database, such as product documents, FAQs, and technical manuals.
Implementation: You can add the relevant data tables to Dify as a knowledge base. The AI can then accurately answer user questions based on this content.