Abstract
Submitting product and listing data to online platforms may involve converting information from various unstructured or disparate sources into predefined schemas, which can be inefficient and prone to error. This disclosure describes a system that can employ a large language model (LLM) agent to serve as an ingestion and reconciliation layer. The agent can receive data from multiple sources and formats, such as text files, conversational inputs, or application programming interface (API) streams. The system is configured to parse this diverse input to identify attributes and to reconcile and merge information pertaining to the same product from the different sources. This method can automate the transformation of unstructured, multi-source data into a structured format suitable for platform submission, which may improve data accuracy and lower the technical barrier for businesses.
Creative Commons License
This work is licensed under a Creative Commons Attribution 4.0 License.
Recommended Citation
Zimmermann, Jonathan, "Generating Structured Product Information from Disparate Data Sources Using a Language Model", Technical Disclosure Commons, ()
https://www.tdcommons.org/dpubs_series/8598