A case study of Generative AI in MSX Sales Copilot: Improving seller productivity with a real-time question-answering system for content recommendation

TL;DR

This paper presents a real-time question-answering system leveraging large language models to enhance sales content recommendation, integrated into Microsoft Dynamics CRM to improve seller productivity during customer interactions.

Contribution

The paper introduces a novel real-time content recommendation system using LLM embeddings, engineered prompts, and a bi-encoder architecture, deployed in a production sales environment.

Findings

System returns relevant content in seconds for large datasets

Successfully integrated into MSX sales Copilot in production

Enhances seller productivity during customer calls

Abstract

In this paper, we design a real-time question-answering system specifically targeted for helping sellers get relevant material/documentation they can share live with their customers or refer to during a call. Taking the Seismic content repository as a relatively large scale example of a diverse dataset of sales material, we demonstrate how LLM embeddings of sellers' queries can be matched with the relevant content. We achieve this by engineering prompts in an elaborate fashion that makes use of the rich set of meta-features available for documents and sellers. Using a bi-encoder with cross-encoder re-ranker architecture, we show how the solution returns the most relevant content recommendations in just a few seconds even for large datasets. Our recommender system is deployed as an AML endpoint for real-time inferencing and has been integrated into a Copilot interface that is now deployed in the production version of the Dynamics CRM, known as MSX, used daily by Microsoft sellers.