B2B
GenAI
Enterprise
2026
Two PoCs designed to keep advisors ahead of every tax change
Two new AI-powered product that monitors legislation and surfaces the most impactful tax news per client, replacing manual tracking so advisors can focus on advice, not monitoring.
The problem
Defining product requirements in a complex, technical space
As the only designer on both products, I worked closely with product managers and tax experts through ongoing conversations and sketching sessions to shape what we were actually building. A big part of my job was helping the team identify what a realistic, shippable MVP looked like — while at the same time sketching out a longer-term vision so that early decisions wouldn't paint us into a corner. The goal was to minimise time to market without building something we'd have to rip apart later.
System Thinking
Designing two products as one coherent experience
Both products live within the same platform and are part of the same proposition, so they needed to feel like they belonged together. I designed them in tandem, establishing consistent UI patterns and flows across both so the experience felt joined up. Because they were owned by two separate product teams, I also acted as the bridge between them — making sure information and decisions were shared, and that nothing got lost in the gap.
AI-powered workflow
Using AI to prototype at speed
For both products I leaned heavily on generative AI tools throughout the process to build interactive PoC prototypes far faster than traditional methods would allow. Being able to put something fully visualised in front of stakeholders early — without committing to it — changed the quality of the conversations around requirements. People react very differently to something they can click through than to a static sketch, and having that fidelity available at concept stage compressed the feedback loop significantly.
Cross-collaboration
Working across product and engineering without losing momentum
I worked across five product managers responsible for different parts of the experience, and supported handover and delivery across two separate scrum teams. With engineering I made a point of keeping them in the loop throughout concept creation — not just at handover. That way feasibility questions got answered early, and we often found simpler or cheaper ways to build things before they were already designed the hard way.
