Skip to main content Skip to Footer

CLIENT CASE STUDY

Accenture in India applies the New now with an innovative digital personal advisor chatbot.

DiPA is a chatbot that leverages artificial intelligence to simplify the HR experience and make it more modern and digital for Generation Y and Z.

OPPORTUNITY FOR INNOVATION

Accenture applies innovation not only for its clients, but also for itself. Accenture in India looked at its challenge of addressing the experience of its many employees who often had to work through a maze of websites, products and services to find and get a wide range of human resources-related information. There was opportunity to create a better experience.

An Accenture internal IT team came up with the idea to develop a digital personal assistant, or chatbot, ultimately named DiPA. The idea was to leverage artificial intelligence to make this experience more modern and digital—and one suited for an employee demographic that is predominantly Generation Y and Z.

FROM IDEA TO SOLUTION

After receiving internal seed funding, the idea for DiPA went through the following steps:

Working Prototype

Working Prototype

Prototype development began with the question: What information is important to employees that can be delivered at their fingertips and conveyed without human intervention?

Scope Definition

Scope Definition

The scope covered the entire employee life cycle—from hire to retire.

These processes were mapped out and employees’ key informational needs for each area were extensively analyzed.

Commonly asked questions-and-answers content related to each of the scope areas was prepared and added to the chatbot knowledge base.

Rapid Innovation

Rapid Innovation

A five-step approach enabled the team to innovate quickly from prototype to production.

The team worked on a technology that was not yet mature and developed a framework to deliver a solution for the enterprise scale, at speed.

Key to the entire process was business sponsorship and involvement from a cross-functional team.

Deploy to Production

Deploy to Production

Once ready, the prototype was deployed into production.

DiPA answered Accenture India employees’ common HR questions by text or voice and quickly became the first level of information support.

Numerous HR-related portals were no longer needed and were decommissioned.

Continued Development

Continued Development

The next phase of the journey is incorporating personalization capabilities.

The internal IT team is developing contextual personalized responses, filtering possibilities and prioritizing them by topics that matter most to employees.

For example, DiPA could provide employees’ leave/vacation balance or suggest an assignment matching their talent profile.

Making Fit Purpose

Making Fit Purpose

Next efforts include making the solution fit for purpose and powering the bot intelligence with Accenture company context, user context, spell check, auto correction/suggestion, content moderation and sentiment analysis of employee feedback.

The ongoing aim is to create an enhanced employee experience in getting HR-related information how they want, when they want.

Ongoing Improvement

Ongoing Improvement

The end-to-end chatbot technology enables DiPA to learn, adapt, improve and grow more intelligent with every interaction.

The next phase of solution maturity is around advising and coaching employees on their career, wellness and learning topics.

BACK-END TECHNOLOGY

DiPA is built on the Microsoft Bot Framework and Microsoft Cognitive Services product stack and hosted in Microsoft Azure.

At the heart of the chatbot solution is Microsoft Language Understanding Intelligent Service (LUIS) and question-and-answer service Microsoft QnA Maker

In the beginning, internal IT was extensively involved with the Microsoft product team as the products were in the early adopter channels and not yet available for enterprise use and scale.

Accenture shared its requirements and feedback, and this active involvement helped enable the maturing of Microsoft Cognitive Services to become scalable.

"We did something innovative and did it quickly with very few dollars. It’s now turning into usage, but just as important, it is informing a go-forward Accenture-wide chatbot strategy that we can leverage, from both business and technology perspectives, having learned where chatbots work and do not work."

SANGITA AGARWAL

Accenture Director – Human Resources Technology