Does anybody has experience with integrating chatbots with Jira Service Desk?

[Radoslaw Cichocki {Deviniti}]

Hi,

I'm interested in hearing what it takes to integrate a chatbot with Jira Service Management Cloud in to save agents' time.

1. What chatbot have you used?
2. What steps need to be taken?
3. How long does it take to set it up and deliver value?
4. What are your results?
5. What is your users' feedback from those chats?
6. What does not work well yet (where is the chatbot you used not mature enough)?
7. Would you recommend it?

Regards,
Radek Cichocki

[Radoslaw Cichocki {Deviniti}]

Thank you, @Bertrand Lafforgue

This is insightful, great answers regarding point 2). I like the overall process that you have described, and what to consider. It's a good chceklist to follow :) 

My remaining questions regard experience from a working implementation, what the results were, what chatbot technology was used in particular, what the user feedback was, etc.

Would you be willing to share your experience in this regard too?

Regards,
Radek Cichocki

[Bertrand Lafforgue]

Hello Radek,

These are very good questions.  Quick feedback:

First, you need to define where your short-term and midterms problems are and not constraint yourself with a solution. What do you need to fix for personas (users, customers, agents,.) in your company?

Once you have a list of use cases that are confirmed by a business owner, you need to define how much value you expect in order to generate an ROI

Then, you need to look at the relevant solution to help you address your problems and could generate the expected ROI

Now, in order to address the juicy use cases in enterprise, you need a platform that can do the following:

-Experience: a user interface available where your users are (Microsoft Teams, Google chat, Web Portal, Slack, Phone/ IVR else) including proactive engagement & notification towards users 

-NLP to understand human language (not keywords) in multilingual when required and generate language 

-Clustering to identify the most common incidents or requests out of your historical data based on semantic 

- Classify based on machine learning to auto-assign tickets to a resolution group, classify a ticket when resolved, etc..

-Search across knowledge bases and filed to retrieve answers 

- Automation to perform automated actions  in your enterprise apps

- Analytics to monitor the performance and the impact of your platform

- etc..

Hope this helps.

Bertrand