Transforming Customer Support with a Unified AI Platform.
My Role
The Team
Solution
What if we put all customer messages in one place and used AI to answer simple questions automatically? That way, support teams could spend more time solving trickier problems. That’s why we built SellSwift — a web and mobile app that helps customer service teams reply faster, stay organised, and focus on what really matters.
Impact of SellSwift
After launching SellSwift, the customer service team streamlined their workflow. With all messages in one place, agents respond twice as fast, provide accurate answers sooner, and handle complex issues stress-free without missed messages or angry reviews.
(This numbers are shared by management)
%
Inquiry Answered
S
Average Response Time
%
Customer Satisfaction
Researching Problem Space
To understand the challenges of multi-channel business communication, I started with secondary research on industry trends and common pain points. I then conducted interviews with customer service teams and business owners to explore their workflows and struggles.
Through my research, I aimed to:
Identifying SellSwift’s target users and their needs
Analysing competitors and their strengths
Exploring AI’s role in improving customer interactions
Discover the goals, needs, motivations, and frustrations of users
Analysing Market & Competitions
To inform my design decisions, a targeted research was done. I deep dive into AI powered CRM business to understand the market standard. Through competitor research I try to understand where we can place ourselves in market of AI powered business communication platform. I found targeted companies offering similar inquiry management or customer service solutions
"Interviews with 10+ customer support agents and business owners revealed the struggle of managing high message volumes across multiple platforms."
We conducted semi-structured interviews to validate our findings and gain deeper insights into how teams handle customer interactions, their pain points, and opportunities for improvement.
Defining Goals
To design an effective solution, I mapped out both business and user needs. We set following goals for SellSwift.
Approaching a Solution
We designed SellSwift to streamline communication through a unified, intelligent platform. Our research showed that agents need a system that reduces friction, automates repetitive tasks, and prioritizes urgent messages.
To achieve this, we leveraged AI-driven automation and smart categorization, ensuring teams can manage conversations effortlessly while maintaining accuracy and speed.
User Flow
To gain a deeper understanding of how users interact with SellSwift, I mapped out their journey from start to finish. This helped me empathize with different scenarios they might encounter, the decisions they need to make, and the paths they take to complete key tasks
Prototype & Testing
With user flows defined, I quickly moved to wireframing, balancing speed and quality for a fast launch. After the first draft, I built prototypes for feedback and internal testing. Iterating through user tests and stakeholder discussions, I refined the designs to align with business goals and user needs.
Final Prototype
Building on the revised wireframes, I created high-fidelity designs and a final prototype. With SellSwift’s branding in place, I focused on integrating its visual identity to ensure a cohesive and polished user experience.
Chatbot insights
Sales Agent Mobile App
What I learn from this project?
Working with limited resources taught me how much I rely on product managers for insights, but it also made me realize the value of direct user feedback.
Simplicity is harder than it seems; getting the design just right took multiple iterations and more time than expected.
Iteration is key; the final design wasn’t perfect from the start, but refining and improving along the way was crucial.
Video Source: Red Dot Innovative