AI/ML Powered Customer Support Ticket Analysis Tool
Case study of how I collaborated with the AI/ML engineering team to design the ML powered customer support (CS) ticket analysis tool, to facilitate product specialists' manual ticket analysis processes, and reduce cost for the company.
Target Users:
Product SpecialistProduct Manager
Design Problem: How to increase CS ticket analysis efficiency and boost confidence of findings.
Design Process & Stakeholders
Design Sprint
Define problem, explore solutions, and identify features with project team stakeholders.
User Interview
Based on the ideas generated, I created a ML diagram mapped with the user workflow low-fidelity mockup. My researcher and I interviewed a dozen target users, and we mapped the feedback back to the diagram.
Wireframing
Then I created the low-fidelity wireframes to engage with my engineers to discuss technical details.
Final Deliverable
The following final design solution captures the end-to-end user experience of the new CS ticket deep dive tool.
1.Case catalog
2.Define ticket elastic search conditions - GUI
3.Define ticket elastic search conditions - CLI
4.Elastic search ticket clusters
5.Elastic search ticket cluster
6.Opportunity sizing - closest ticket distance
7.Opportunity sizing - increase ticket relevance distance
8.Opportunity sizing - estimate opportunity size
9.Modify ticket search criteria ad-hoc
10.Flag irrelevant ticket - provide feedback data to machine
11.Use above bar charts as filter selection
12.Report - annotation
13.Report - edit annotation
14.Report - review final ticket list
