Overview
In 2024, a leading global e-commerce platform commissioned us to imagine what its service could look like if designed around AI from the ground up. The brief was ambitious: challenge existing conventions, think beyond current technical limitations, and absolutelyavoid retroffiting a chatbot next to their existing interface.
I led a multidisciplinary team of four specialists, combining expertise in AI, business, technology, and service design. Over a few focused sprints, we developed a bold vision supported by interactive prototypes.
- Role
- Team lead and lead designer
- Timeline
- Start of 2024
Project highlights
Running a business on the e-commerce platform requires a team looking at logistics , marketing, finance, etc. We envisioned an agentic experience as the backbone for businesses with dedicated agents supporting your operations 24/7. Specialised agents would be in sync to help you to, for example, list a product in the platform, analyse similar items to help you differentiate, and help you keep running your shop through a dashboard with proactive recommendations.
Vision & strategic design
- Leading a multidisciplinary team of four specialists
- Defining an AI-native experience beyond traditional interfaces
- Using scenario-driven storytelling to communicate the vision
Prototyping & stakeholder engagement
- Presenting concepts and technical demos directly to the Head of Design
- Combining Figma experiences with working AI prototypes in Streamlit
- Working in fast-paced weekly sprints
The challenge
The platform supported a vast ecosystem of businesses, yet many everyday tasks required navigating complex interfaces and filling out lengthy forms. Running an online business also involves managing products, logistics, marketing, and finances, often across disconnected workflows. Simply adding AI features to the existing experience would not solve these underlying problems.
The opportunity: What if the platform could actively help businesses run their operations instead of making them manage every task manually?
Reimagining commerce around AI agents
We envisioned an AI-native experience built around a network of specialised agents working together to support businesses around the clock.
Instead of navigating complex menus and forms, sellers could work with agents that understood their business, coordinated tasks, and proactively suggested next steps. For example, when listing a new product, agents could help sellers create the listing, analyse competing products, identify opportunities to differentiate, and recommend improvements. A central dashboard would bring these activities together, giving business owners a clear view of their operations and timely recommendations.
The goal was not just to simplify individual tasks, but to shift the platform from a tool businesses operate to a service that actively supports them.
From vision to working prototypes
With only a few sprints to explore a broad ambition, we focused on making ideas tangible early. We worked in weekly cycles, combining opinionated design directions with scenario-based storytelling and working AI demonstrations. Figma helped us communicate the end-to-end experience, while Streamlit allowed us to demonstrate real AI capabilities behind selected interactions.
Each presentation to the Head of Design gave us an opportunity to test assumptions, gather feedback, and refine or pivot our direction.
This approach helped turn an ambitious AI vision into something stakeholders could see, experience, and discuss.
The outcome
We delivered a clear AI-native experience vision, supported by interactive prototypes and concrete business scenarios.
The package gave the Head of Design material to bring into C-level discussions and a starting point for internal teams to explore future product directions. The work received highly positive feedback and demonstrated how an agent-based approach could fundamentally change the experience of running an online business.