Hi YuYong, It's very exhilarating to receive your interest in time series database builtin language AI capabilities. In fact, we've been activily exploring the builtin time series forecasting capabilites within the past few quarters through the development of AINode, as time series data are the dialect of time series databases. On the other hand, the natural language is natively more closer to relational databases. Thus, although we left `chat` inference interface in AINode, we do not introduce the LLM capabilities currently. As a result, we warmly welcome a discussion about how LLM can benefit our time series area. Looking forward to your research survey!
Sincerely, Yongzao On 2025/12/26 09:01:55 yong yu wrote: > Hello everyone, > I've noticed several databases[1] recently introducing query capabilities > related to large language models (LLMs), such as embedding generation, text > analysis, and AI-driven data processing. This trend is intriguing and holds > potential value for IoTDB users as well. > > I'd like to gauge interest in implementing similar AI features within IoTDB. > If demand is strong enough, I'm willing to implement some AI capabilities via > User-Defined Functions (UDFs). > > I look forward to hearing your thoughts on this idea: Do you find such query > functions practical? Which AI capabilities would be most valuable to IoTDB > users? > > > > [1] https://doris.apache.org/docs/dev/ai/ai-function-overview/ > > Sincerely, > > YuYong > > > >
