Tars Debug Mode
Company
Tars
Role
UI/UX Designer
Team
Vinit Agrawal, PM & CTO Vaibhav Kubre, Developer
Date
October 2022
The requirement
"The AI chatbots/assistants created via Tars are often quite complex, with heavy API integrations and lengthy conversation flows. If there is an error or break in the connections, it gets extremely difficult and time-consuming to comb through the chatbot flow to find it." This was a pain point which was brought forth by the customer success team at Tars. Recognizing the importance of having such a feature which could solve this porble, the product brainstormed on ideas on how we could tackle this pain point. This resulted in the conception of the Debug Mode feature. I handled the design and overall experience of the feature based on calls with the customer success team.
I went to the drawing board (or rather the Figma canvas) and based on the research material, conceptualized the first version of the feature. Both the initial iterations were really tech-perspective heavy.
The final design
Since the release of this project, the customers are now able to test their chatbots without reaching out to the Tars' Solution team. This in turn has also reduced the solution team's load of work. In addition to this, the time taken to find out errors in the conversation flow and correct them is now a fraction of what it used to be earlier. To give you a glipmse of how this feature was received in the team, you can have a look at this conversation snippet from the Tars Slack workspace.
