Cognitive Helpdesk

While we were working on creating a better usable tool for our help desk agents, we noticed that there was a significant pain point that was hindering the productivity of our agents when dealing with service tickets that were simple to solve yet very redundant. There was an inflow of these kinds of service tickets on high volume which left them with little time to handle complex issues.

We gathered our help desk agents and conducted several user interviews to deep dive into their pain points and how these pain points were affecting their overall productivity. We understood their demographics, their goals at workplace, their motivations, and frustrations.

We discovered that a lot of manpower was getting spent in addressing mundane and simple issues, most of which could be automated. This same manpower can then be used better in scenarios where a human intervention was required to solved more complex issues.

We also conducted user interviews with the counterpart persona who were the users who create tickets. There were certain pain points that prevented users from creating a ticket quickly or skip it entirely. They had to fill long forms even for simple and straight forward issues which was time consuming and error prone.

We ideated on different ways to solve this problem for both the personas. We brainstormed several ideas before we stumbled upon using cognitive technology that can replace a human help desk agent for mundane and repetitive issues, so the manpower can be used to solve more complex issues. The envisioned cognitive helpdesk can act like a virtual agent that converses with the users, assisting them in raising tickets in a more natural way than typing and filling long forms.

We identified all the scenarios where cognitive helpdesk can solve the problems and designed the following features:

• Automation of low complex, repetitive helpdesk tasks such as Password Reset, status inquiry, etc.
• Personalized conversational interface
• Integration with enterprise business applications
• Auto ticket logging, triaging and resolution
• Knowledge base driven user’s query addressal
• Contextual troubleshooting & self-help instructions to users