Solution
Trustpilot is a global consumer review platform used by over 63 million people each month. TrustLayer is Trustpilot's data product for an AI-driven world. It reimagines how people interact with Trustpilot beyond the website, meeting them in the platforms they already use and surfacing review insights at the moments decisions are made, from LLMs and social media to the tools consultancies and investors depend on.
Role
Senior Product Designer
I co-led design sprints and I pitched early concepts to senior stakeholders, shaping the vision for TrustLayer. I created partner-focused prototypes showing how Trustpilot’s data could integrate across platforms. This work secured partnerships, informed its launch at HumanX, the world’s largest AI conference, and set the stage for a dedicated product squad.
Activities
Design sprint facilitation
Product Vision
AI-augmented prototyping
Concept pitching
Rapid concept generation
Research synthesis
User research and interviews
UX design
Stakeholder alignment
UI and visual design
Team
Senior UX Researcher, Senior Product Manager, Engineering Manager, Senior Data Scientist
Duration
October 2024 - January 2025
1 month sprinting, 2 months developing
Impact
Proved market value
Secured live trials with hedge funds and global consultancies to test TrustLayer in real-world use.
Opened doors with big tech
Initiated licensing conversations with major tech platforms
Sprint Cycle 1
I planned the sprint as a compressed 5-day flow to move quickly under time pressure. Day one was about gathering diverse perspectives, while the rest of the week was carried forward by the core team to shape, test, and refine ideas before sharing them back with stakeholders at the end of the week.
After the workshop, we focused on developing the ideas with the most potential. I generated quick prototypes in Lovable AI so users had something tangible to react to, making feedback more meaningful. As a team, we tested these concepts to give us insight into what people valued, what they didn’t, and how concepts could be improved. To ground that feedback in reality, we assessed each concept using a feasibility and viability framework, which clarified what was achievable in the short term. We each pitched a concept to senior stakeholders and their input helped us prioritise focus areas for the next sprint.
We ran this sprint format across two more weeks, generating, refining, and pitching ideas for both businesses and consumers. Many clustered around three patterns: integration into existing platforms, community trust, and personalised discovery. These directions were useful, but they felt like safe, incremental extensions of Trustpilot.
Sprint Cycle 2
I felt that the ideas we were generating were safe. I reached out to the wider UX team for processes we could use that would provoke more radical thinking, and Atlassian’s Disrupt framework was suggested. I created bespoke constraint cards and problem statements and we spent three intensive days using the Disrupt method to generate unexpected ideas.
🐝
Status Quo Bias
We tend to not change an established behaviour (unless the incentive to change is compelling)
📊
Trust Without Data
Design as though you have no access to data, reviews, or ratings. You only have real time interactions
🌫️
Everything is Ephermeral
Assume all information on the platform is temporary and will disappear after a short period
💻
No Screen
In a voice-only environment how do we communicate what is trustworthy?
🔋
We are an Energy Drink
We imagine Trustpilot as a burst of clarity and confidence, instantly energising for decision-making.
🔄
B2B & B2C Switch
We allow consumers to rate businesses but businesses can now rate consumers. What would emerge?
🤷♀️
Need for Certainty
We crave certainty and are more likely to take action if specific information is available
⚓
Anchoring & Adjustment
When making decisions we rely too heavily on one piece of information which is often irrelevant
This format unlocked a new level of creativity: over three days we generated more than 100 ideas. We then ranked them by boldness and viability, before pitching the strongest concepts back to senior stakeholders. I analysed the concepts we generated and they fell into four key themes
Embedded Trust
Trust should meet people where they are and embed into platforms they already use.
Proactive Guidance
Trust should actively shape decisions in real time, not just sit as a static score.
Power of the Collective
Communities and trusted individuals carry more weight than faceless averages.
Transparency in Action
People want to see how businesses really behave - their responses and values.
Each of these ideas pointed to a bigger shift: trust shouldn’t be something you search for, but something that’s already there when you need it. That realisation reframed our direction. What if trust could move fluidly, like a layer of data running through the web, surfacing wherever decisions are made? That question became the seed of TrustLayer - a system designed to bring verified human voices into any platform that shapes choice, at exactly the right moment.
Prototypes
With a direction in place, the next step was to show what it could look like in practice. I created prototypes for senior stakeholders and partnership meetings to illustrate how TrustLayer might appear in real-world contexts. These visuals helped potential collaborators imagine how they could integrate Trustpilot data into their own products and workflows.
TrustLayer for AI Assistants • As people turn to AI tools for everyday decisions, TrustLayer brings verified human voices into the conversation. When a user asks for recommendations, it draws from real Trustpilot reviews to ground answers in genuine experiences, making AI-generated suggestions more reliable.
TrustLayer for Social Platforms • On platforms like TikTok, people make quick decisions in the moment. TrustLayer brings real reviews into those spaces, so whether someone’s browsing TikTok Shop or watching a live stream, they can see genuine feedback at the right moment before they buy.
HumanX
These explorations came together in the public launch of TrustLayer at HumanX, where Trustpilot shared its vision for building trust in the age of AI. The keynote showed how real human voices could appear across different platforms, along with ideas like a sentiment analysis dashboard that revealed new ways our data could help people make more confident decisions. Seeing the work come to life on stage made the impact of our explorations feel real.
Learnings
This project began in deep ambiguity, with shifting ownership, unclear priorities, and a hard deadline for HumanX. Working at that pace alongside my full-time role taught me how to bring purpose to uncertainty and move forward without waiting for perfect clarity. Early on, it was difficult to build momentum with a revolving cast of stakeholders, which meant ideas were restarting each sprint instead of evolving and building on previous work. Since then in my work outside of design sprints, I’ve made it a habit to create a one-page RACI and maintain a visible decision log - so direction always remains consistent even as teams evolve. In many ways, the chaos was the point. It taught me how to find direction in uncertainty and trust my instincts as a designer when nothing feels certain. After shaping what trust could look like in the future, I wanted to turn my focus to how people understand it today.