Designing a smarter GenAI workflow for AI developers
Three iterations, 32 international users, and a product that stopped being a separately-named add-on along the way.
- Role
- Product Designer
- Client
- NDA - Global Tech Enterprise
- Date
- Jul 2024 – Oct 2024 (4 months)
- Scope
- UX research, product design, product strategy contribution
- Tools
Figma
Miro
TL;DR
Brief was vague on users. A separately-named GenAI add-on, scoped to launch in a planned marketplace alongside an existing enterprise AI platform. Three predefined personas, no evidence on which was primary, and an undecided third-party integration.
Research shaped the product, not just the interface. 32 international data scientists and AI developers across 6 interviews, 12 usability tests, focus groups, and an in-person playtest. 12 key insights emerged.
The add-on became a premium feature, not a separate product. From a planned marketplace launch to a premium feature embedded in the platform's core project creation flow.
Context
A separately-named product, wrong shape of feature.
An enterprise AI platform wanted to add GenAI capabilities. The initial spec framed this as a separately-named, separately-priced add-on, scoped to launch in a planned marketplace alongside the platform. Extensive on the technical side but vague on users: three predefined personas, no evidence on which was primary, and an undecided third-party integration. 12 weeks to figure it out. My scope: research, persona narrowing, user flows, and design across three iterations to a final product vision. Engineering, hosting, and infrastructure sat with other teams.
The product definition shift
The most valuable design work wasn't on the interface.
Through 6 user interviews, 12 usability tests, focus groups, and an in-person playtest with the client's AI scientists, we narrowed to the AI developer as the primary user. 32 international data scientists and developers spoke with us across the project. Access was a problem initially. Once in front of them, insights came fast.
Twelve key insights shaped the direction. Users wanted full customisation, not just quickstart options. They expected AI assistance as a baseline. Bundling features created confusion, not simplicity. The third-party tool wrapped into the experience had its own brand, and most users didn't recognise it. We removed the brand references and surfaced the feature by what it did for the user.
The shift: the product stopped being a separately-named add-on. It became a renamed premium feature of the core platform, embedded in the project creation workflow.
Iteration discipline
Three rounds, none of which got it right on the first pass.
Three design proposals, each shaped by testing and feedback.
The first split project creation into ML vs GenAI paths upfront. Users liked the simplicity, but didn't like deciding that early.
The second introduced templates with customisation. Templates landed much better, though users weren't always sure if a template was local or cloud-based.
The third integrated the third-party evaluation tool as an entry point. Users found it confusing (the unfamiliar brand compounded it).
The final took the best of each: template selection with full customisation, a streamlined setup panel, the evaluation tool integrated without exposing its branding, and a redesigned project landing page focused on experiment runs.
Outcome
A changed product direction and a complete product vision, handed off with a clear roadmap.
Five core screens delivered (template selection, customised template setup, custom project creation, premium trial modal, experiment runs) with a roadmap covering feasibility, development, and future features.
32
users
12
weeks
3
iterations
12
key insights
What I took from it
Sometimes the most valuable design work isn't about the interface.
It's about helping a team figure out what the product actually is before building it. The brief framed this as 'design an add-on'. 32 developers later, the more useful framing was 'embed a premium feature in the existing product'.
