IT Best Practices

Natural Language Processing and Machine Learning Radically Transform How the VMware Business Interacts with Business Systems

by: Sunder Vudhya, VMware IT Director, Digital Order Ops and Advanced Technologies and Balaji Parthasarathy, VMware Senior Director, Enterprise Architecture and Experience Design

VMware IT has increased its application footprint to deliver solutions for a variety of business needs, resulting in a distributed system that meets both domain needs and business goals. A distributed system approach, however, brings its own challenges because the data generated in the system resides in many different places.

The Business Operations team receives 120+ inquiries each workday–or more than 10,000 inquiries per quarter–about the status of the orders, opportunities, quotes, PO, invoices, renewals, etc. These inquiries arrive from various sources, including partners and distributors. Operations analysts must gather data from various Systems of Record, including sales order bookings, sales finance, geo sales, revenue, and work with other teams to respond to these inquiries. It can be time-consuming for the analysts to retrieve and validate the information for data quality and accuracy because they must access multiple systems to retrieve it. Each inquiry is manually tracked as a separate case. Depending on the number of systems involved, business groups may even reach out to IT to get additional information.

Our challenge was to create a self-service model that could automatically retrieve information and answer simple requests while maintaining a high level of responsiveness, quality, and customer satisfaction. A streamlined process would ensure faster responses and enable analysts to focus on more complex issues.

Automating a Bot Solution
IT developed next-generation, a Natural Language Processing (NLP) platform to increase the responsiveness of some tasks. It enables Bot developers to train machine learning (ML) models for intent classification and entity extraction. vAssist is available as a SaaS model and exposes easy-to-use REST APIs to train and parse natural language inputs. It also provides a dashboard where developers can annotate and visualize data expressions and ML models in the cloud. We also integrated with one of VMware’s chat platforms. Users can make service requests using the chat platform.

Architecture and Implementation

VMware OM Assist architecture

OM Assist (also called Business Assist) is a user-friendly, simple-to-use, and bot-built software. It is implemented on the chat platform using concepts. It works like this. When logged into the VMware network, an end user can search for OM Assist in the chat communication channel, then ask for the available options.

An intuitive conversation is initiated by the bot, which provides a list of services that are available, such as quarter-end close data for executive leaders; hold reports for various roles; and the status of sales orders or service contracts based on the PO number, quote number, customer number, etc. Once the request is made, OM Assist responds with real-time data. The user does not have to remember the options each time or wait for an analyst to track down the requested data.

OM Assist was developed using a library of APIs (microservices) that cater to the needs of different business units. An API using AI technology enables the user to self-serve their inquiries. Users can access the data when they need it, including reviewing the status of the data or creating business operational reports in real-time, depending on the business use case. Services are accessed based on individual roles and responsibilities.

VMware OM Assist Image

The Results
The biggest advantage of OM Assist is its self-service model, which delivers automated information 24×7 on the status of the orders, opportunities, quotes, POs, invoices, and renewals. The streamlined business process provides an integrated experience for users without any human intervention. By using emerging technologies, IT has created a scalable process that enables Business Operations to handle more inquiries and focus analyst efforts on resolving more complex requests without sacrificing responsiveness, quality, and customer satisfaction.

Since its adoption two years ago, we have enabled more than 23 services that are critical to business users and achieved a reduction of ticket avoidance by ~10%. The results will be even more amazing when we extend OM Assist to individual contributors within Business Operations. Moreover, IT is planning to introduce 50+ new features covering a wide range of business use cases.

VMware on VMware blogs are written by IT subject matter experts sharing stories about our digital transformation using VMware products and services in a global production environment. Contact your sales rep or to schedule a briefing on this topic. Visit the VMware on VMware microsite and follow us on Twitter.   


3 comments have been added so far

  1. Great Sunder !!!. Very nicely documented and explained the OMAssist and it’s back bone. It is getting expanded across BUs operations now !!.

  2. This platform has now become such an integral part of business, it seem hard to comprehend operating without this interactive bot.
    With such capable and effective team, VMware juggernaut is still in its early stages.


Leave a Reply

Your email address will not be published. Required fields are marked *