This content originally appeared on DEV Community and was authored by Chen Debra
DolphinScheduler is planning to introduce a Copilot-style chat assistant!
Powered by large language models (like OpenAI), it will help users write SQL intelligently, optimize code, answer questions, and greatly improve the development experience.
DolphinScheduler is keeping up with the wave of AI transformation. Staying current isn’t just a slogan—it’s about taking action!
This fantastic idea is still in the design phase. Whether it gets added to the official DolphinScheduler roadmap depends on the power of our contributors!
Here’s the current vision from the initiator of this DSIP—see if it resonates with your thoughts:
Motivation
As large language models grow more powerful, the quality and accuracy of SQL generation has reached a new level.
Now is the perfect time to introduce LLMs into scheduling systems to help improve development efficiency, optimize code, and answer questions users encounter during usage.
DS Copilot is a data-intelligent assistant built on third-party LLMs. It will integrate metadata from DolphinScheduler’s data sources and assist users in writing higher-quality, standardized SQL tasks.
Design Details
Add a Chat Module
This module will receive user input → enrich it with metadata from the current SQL data source → generate optimization suggestions and send them to the LLM → then return and display the response in a chat window for the user.
LLM Registration
This feature allows registration of LLMs. The initial implementation will support OpenAI, with plans to expand to other models in the future.
Join Us!
If you have valuable suggestions or practical experience, you’re welcome to comment on DSIP #17334 and join the discussion GitHub Issue 17334:
What do you hope Copilot can help you with?
Which large language model do you think fits best?
Do you have any design ideas or implementation suggestions?
Let’s brainstorm together and drive DolphinScheduler’s intelligent evolution forward!
This content originally appeared on DEV Community and was authored by Chen Debra