Kor: Unleashing the Power of LLMs for Structured Data Extraction

Kor is a prototype tool that uses LLMs to extract structured data from text. It is slow and prone to crashing for long pieces of text, but as LLMs improve, these issues may be mitigated. It is best suited for extracting data from text that matches an extraction schema and providing natural language access to an existing API. Users discuss alternative models and use cases, such as generating web scrapers and data processing steps on the fly, and extracting metadata from HTML. Another tool like Kor is Marvin.

#AI #LLM #GPT4 #Prompt #GPT #OpenAI