Employee Virtual Assistant

All Companies value employee engagement very much. What employees see, hear, and feel in their daily job is a part of the employee experience. A helpful and friendly chatbot can improve the efficiency of internal communication.

My client company set up a goal with the employee experience improvement. Creating a Self-service Center and a Virtual Assistant for employees were the solution to address the team bandwidth issue. I was hired as a Conversation Designer to contribute to the entire user journey for the Employee Virtual Assistant (or chatbot as we called it) UX research, framing, and UI design.

Beyond fully functional, the chatbot needs to be well-spoken and speak in grammatically correct sentences. The chatbot also needs to have its personality. However, personality does not shine through individual chat messages. The personality and an incredible user journey are a sum of the way that the bot makes a user feel at the end of a complete conversation. Therefore, the UX design with a holistic view is essential in this project.

The final goal was to enable the employees' self-service ability and improve the employee happiness score.

As a conversation designer who understands the bot's capabilities the most, I needed to provide a framework for the bot training. ML algorithms need LARGE amounts of data. The bot needs to identify which user utterances lead to which journeys and help the developer understand this correlation.

I mapped out the employee life cycle and user journey map. Visualizing these essential touchpoints helped us have a better understanding of the user journey.

We also sent out surveys to employees to recognize the biggest pain point and questions.

IT and HR are two primary departments to deliver and steward company culture and employee experience. I collaborated with stakeholders from these teams to build a joint vision and strategy.

I was responsible for ensuring that the chatbot could solve problems and employee questions most efficiently and be humanly possible.

In my design brief, I set up a couple of key points for the chatbot:
- What we can do (Tech limits)
- What we want to do (Expectations)
- Tone (Personality)
- Logic (Question database)
- Scripts
- Monitor performance
- Educate users

It was my task to find the best structure for a conversation. To do this, I needed to understand the mindset of the user. What kind of questions do people have, what motivates them, and what are their concerns and anxieties? Structured information in a way that people feel understood, get their questions answered, and where the conversation moves forward engagingly was my goal.

Here we're my process:
- Built a bot persona
- User Cases
- Sample Dialogue and role play
- Prototype (Figma and Voiceflow)
- Usability Test

I also kept in mind that to maintain a short and precise copy. The clickable and actionable elements with the bot also need to ensure optimum user engagement.

According to the company's reimbursement policy, I created this sample dialogue:

Tracking metrics like query completion and automation levels told us exactly how much or how little we were helping users. The proof is always in the conversation rate.

I think conversation design is a great field to explore for myself since I love UX, writing, logical thinking, and Product Design.