ADP, a global provider of cloud-based Human Capital Management solutions, engaged us to create a proof of concept using the latest AI technologies to improve the experience of their clients and associates by providing tools and technologies to ‘Globalview Associates’ to get answers quick and easy.

Think of these guys as the ‘support’ team of ADP that handle client (your HR team’s) requests.


My Role

I led the entire design of the product/experience - GUI, VUI and Multimodal explorations. 

Product Strategy - Product Design

1732 - 1082


Current Situation

Currently, Globalview provides a series of ‘technical assets’ for associates to assist clients as calls and tickets are submitted. These systems include ADPWorks, SharePoint, SAP CRM and various file shares. In addition, associates depend on each other in certain scenarios to get answers from peers. While answers are always discovered and communicated with the client, the amount of time it takes to provide these answers and number of steps, in certain cases, result in a less that exceptional client experience. 

To comply with my non-disclosure agreement, I have removed and obfuscated confidential information in this work. The information here is my own and does not necessarily reflect the views of anyone else.



  • Seeking ways to leverage the latest in AI to create great client experiences

  • Transform manual ADP forms to interactive Smart Forms.

  • Globalview anticipates significant client growth and is looking for ways to sure the growth without massive scaling of the support team

  • Improve client experience by providing tools to associates to get the best answers quickly


Business Outcomes

While this proof of concept was technical in nature, the focus was on the desired business outcomes ADP was hoping to achieve in deploying the Intelligent Assistant. 

These objectives include:

Improved productivity: Reduced internal handoffs and improved workflows.

Associate Adoption: Familiarity with modern digital conversations.

Associate Retention: Empowered to be self-sufficient and capable of solving problems quickly resulting in increased morale.

Greater Insights and Analytics along the client journey. i.e. new questions and insights into what “other” actually means. 

Creates Positive Experience between Associate and Client

Scalable across ADP: using repeatable process, templates and patterns. (horizontally and vertically within ADP)

Faster Response: Client and associate

Improved Client Experience: Leading to improved client retention



We developed user stores and personas to help guide our VUI/GUI journey taking cue from Google’s conversational design guidelines and Amazon Alexa Designing a Voice Experience. This allowed us to focus more on a user problem and how to navigate through this experience. 

To help define the journey we created same dialogs revolving the persona and their particular needs.

To comply with my non-disclosure agreement, I have removed and obfuscated confidential information in this work. The information here is my own and does not necessarily reflect the views of anyone else.


Concepts & Testing

Language, comprehension and monitoring usage were essential for the success of this product. We iterated on different concepts and validated using multiple solution/answer providing platforms before landing on a final concept.


User Testing

We used a combination of remote quantitative and qualitative moderated tests.

The latter helped provide insights on the overall idea and behavior of how users would react with particular flow, messaging and monitoring of their usage during an event.

We validated different solution/answer representations using Usability Hub to make sure our assumptions and the experience were comprehensive.

Hyundai - Response - Final - Delivered - Persona.001.png

What I was trying to solve for 

I proposed the creation of an Experience as a Service (EaaS) architecture that demonstrates the efficiencies that can be gained by creating a simple conversational user experience (GUI + VUI) to get questions answered quickly. The solution consisted of several components working together to accept questions, collect answers, prioritize answers and display solutions. The POC demonstrated the solution’s ability to navigate an associate through a relatively complex scenario by asking clarifying questions that refine the results and reduce the possibility of misleading answers. 

In addition, the intelligent assistant was designed with scalability in mind, both within the Global View team, as well as across other areas of ADP;  allowing ADP to leverage the investment across multiple lines of business. 

My objective was to leverage artificial intelligence and machine learning (with the help of Alexa) to get the right information to the right person, in less time, while creating an amazing client experience.

Artboard Copy.png
Artboard Copy 2.png
Artboard Copy 5.png


I tried to extend the EaaS and create a multimodal experience on Amazon Alexa. I created the conversation flows on a tool I often use for voice prototypes for Alexa and Google - VoiceFlow. Then started prototyping the experience just to sell the ‘concept’ to the client.

The biggest challenge I faced was that if the experience was solely on VUI (without any help of GUI) since the answers/solutions were in so many steps with so much information, there was a huge chance that the associate won’t be able to follow up with the directions and get confused, lost and eventually frustrated. I tried to over come this by giving Alexa “breathing space” and after each step try and get confirmation from the associate to either “shall I continue?”, “tell me when you are ready.”, “Do you want me to repeat that for you”, etc.

I used the same research material for the bases of Persona building and conversation flows keeping the change of intentions, context of use, and possible change of action in mind. The overall goal was to:

  • Craft conversations that are natural and intuitive for the user to understand and Alexa to take appropriate actions on

  • Create a multimodal experience to Scale my conversations across all devices to help users wherever they are in whatever context they are

To comply with my non-disclosure agreement, I have removed and obfuscated confidential information in this work. The information here is my own and does not necessarily reflect the views of anyone else.