Smart Online Support

Project Overview:

Context:
Azki is the leading insurance aggregator in Iran.
We help users compare prices, understand what they buy, and complete insurance fully online.

Story:
The support experience was fragmented and confusing — users didn’t know where to go for help, leading to frustration and unnecessary calls.
The goal was to design a flow-based and smart self-service system that helps users resolve issues independently before contacting an agent.
This included clear category flows, contextual actions, and an AI-powered assistant trained on real FAQs and support conversations.

My Role:
Product Designer — End-to-End (Research, UX, UI, Data, Testing, Handoff)

Platform:
Web + Mobile Web

Timeline:
6 months (3 design — 3 development)

Team:
PM, Dev Team, QA, EDM, Data Analyst, UX Writer, Visual Designer, Design System Manager

User Segment:
All Azki users who need consultation, order tracking, or problem reporting

Business Goals:
– Reduce support calls and ticket load
– Improve customer wait time and satisfaction
– Build scalable structure for AI-driven support
– Reduce operational costs in customer service

Problem:
– Users were lost in the old support flow — no clear entry points or hierarchy.
– Most users directly called support for simple issues that could be automated or self-solved.

Constraints:
– Follow current APIs and user flows
– Design system and technical limitations
– Alignment with operation and call center teams for unified tone and content

Research Methods:
– Usability testing (Redesigned version — before launch)
– Analytics and funnel analysis (Metabase)
– Customer support logs review (Chat + Call Center)
– Stakeholder interviews
– Competitor benchmarking
– Heuristic evaluation

Solution:
We designed a flow-based support system with structured categories and contextual actions.
An AI-powered chatbot (RAG + Redis memory) handled FAQs, routing complex cases to human agents.
The design emphasized clarity, scalability, and seamless transitions between automation and manual support.

Outcome:
−15% support calls
−50% customer wait time
−30% new operations hiring costs
Launched Azki’s first self-service online support, reducing manual workload and improving user experience.

Private Case Study

This project is protected under NDA. Please email me if you’d like to request access

Let’s Connect 🤙🏻