Challenge
Teams had to search across a large volume of PDF, Excel, and Word policy documents, and early RAG attempts could not reliably answer questions that depended on table-based rules and exceptions.
SoftBank's February 2026 public case shows how structured RAG on top of internal regulations and rule documents turned a hard-to-use policy search flow into a daily operational tool.
uses in the first 2.5 months
work time created
regulations in the managed rule set
Each case study page keeps the narrative tight: the constraint, the intervention, and the measurable effect.
Challenge
Teams had to search across a large volume of PDF, Excel, and Word policy documents, and early RAG attempts could not reliably answer questions that depended on table-based rules and exceptions.
Solution
SoftBank combined data structuring with a RAG workflow and connected it to ChatGPT Enterprise in a controlled environment. By converting policy and spreadsheet content into AI-readable structure, the team improved answer quality and made rules lookup practical for everyday work.
Public source
This case is based on a public disclosure rather than a confidential client delivery.
We can translate the same structure into your inquiry flow, knowledge system, or automation roadmap.