Abstract
A cross-team enterprise chat system comes with a set of challenges, not typically addressed in an out of the box solution. A large enterprise tends to have hundreds of thousands of unstructured documents that continues to grow exponentially in number over time. A standard RAG system or a traditional Question-Answer Mapping (QA) AI agent built on top of such a corpus would be impractical and inefficient as it would need to comb through the entire knowledge base to fetch relevant data to the user’s query. The existence of data in multiple silos with different teams further complicates the issue, as typically teams may be reluctant, or would have to go through complex compliance processes to share their data. Thus, scalability, Privacy Awareness, and resource efficiency are the biggest challenges to be addressed in any methodology that aims to improve this entire process. To handle these challenges, we propose a practical, scalable, privacy aware distributed solution designed in such a way that the individual teams retain maximum control over proprietary data while the overall system is made more efficient by avoiding redundancy, through grouping similar questions together, tagging them by topic and storing the relevant chunk positions along with them.
Creative Commons License
This work is licensed under a Creative Commons Attribution-Share Alike 4.0 License.
Recommended Citation
INC, HP, "An AI Enabled Enterprise Question-Answering (QA) System Optimized for Scalability, Privacy Awareness and Efficiency", Technical Disclosure Commons, (March 30, 2026)
https://www.tdcommons.org/dpubs_series/9653