I worked on the design of the eSIM experience as part of SMARTY’s introduction of this capability into its product offering.
The work focused on translating a technically complex and evolving solution into a usable and coherent customer journey, ensuring that the experience remained intuitive despite several system-level constraints.
The design process was highly iterative. Before moving into design exploration, I mapped out the existing journeys end-to-end to build a clear understanding of how current flows operated and where the new eSIM experience would need to integrate. This included defining both current and proposed process flows early on, which helped establish a structured foundation before moving into low-fidelity design.
Multiple low-fidelity journey variations were then explored, as new technical limitations were progressively uncovered by architects and engineering teams becoming more familiar with the technology. This required continuous adaptation of the experience, balancing user expectations with feasibility.
A key challenge emerged around the inability of devices to reliably confirm when an eSIM had been successfully installed. This created a direct conflict with established usability principles, particularly around visibility of system status. To address this, alternative feedback mechanisms were designed to provide users with reassurance and clarity at critical moments in the journey.
The work was informed by a combination of exploratory, behavioural, and comparative research. Given the novelty of the capability, there was limited internal data available, which required a stronger reliance on external benchmarking and internal behavioural analysis.
I collaborated with the Funnel Improvements team and used DataDog to identify key drop-off points and bottlenecks across the existing purchasing and onboarding journeys. This helped surface areas of friction that could be addressed as part of the eSIM rollout, ensuring that the work not only introduced a new capability but also improved the overall experience.
A detailed competitive review was conducted across multiple MVNOs and telecom providers to understand how similar constraints were being handled in market. This helped validate that certain limitations, such as the lack of system feedback during installation, were not isolated issues but industry-wide challenges.
Alongside this, existing customer journeys within SMARTY were analysed, particularly around onboarding, checkout, and number transfer flows. This allowed for identification of friction points and opportunities for simplification, especially within PAC and STAC scenarios.
To support synthesis at scale, I cautiously used ChatGPT under my supervision to help consolidate large volumes of research inputs, improving speed of analysis while maintaining accuracy through manual validation.
The research also involved close collaboration with technical stakeholders, where constraints were continuously uncovered and translated into design considerations. This created an ongoing feedback loop between research, design, and engineering, ensuring that decisions were grounded in both user needs and technical reality.
The introduction of eSIM had a significant impact across the entire purchasing lifecycle. Changes were required across the homepage, plan selection, account creation, checkout, legal content, and order confirmation, extending beyond the core product journey into email and SMS communications.
As part of this, existing flows such as PAC and STAC were revisited and refined. Opportunities were identified to simplify and consolidate previously fragmented scenarios, improving clarity and reducing cognitive load for users navigating these processes.
The experience differed between web and mobile app, particularly where the journey required interruptions or transitions mid-flow. These differences introduced additional complexity, requiring several iterations to ensure consistency while still respecting platform-specific behaviours.
At the time, the existing design system was planned for decommissioning, with a new system being developed but not yet ready for adoption. This created a transitional phase where the new eSIM journeys needed to integrate with the legacy system while also anticipating future changes.
To address this, I designed a set of temporary components to support the new experience, ensuring consistency across the journeys while filling gaps in the current system. These components were created with forward compatibility in mind, allowing for smoother migration into the new design system once it became available.
The final outcome was a cohesive and adaptable eSIM experience that aligned technical constraints with user expectations. The work also helped improve legacy flows and established a stronger foundation for future iterations as the capability matured.
I led the research and design of improvements to the login and authentication experience, focusing on reducing friction across key journeys such as login, account recovery, and account access.
The project combined extensive research with rapid design delivery under significantly compressed timelines.
Originally planned as a longer engagement, the project timeline was unexpectedly reduced from two months to two weeks. This required an accelerated design process, including intensive working sessions to ensure delivery within the new timeframe.
A high-fidelity set of designs was produced to represent the ideal authentication experience across multiple scenarios. These included login, registration, password and email recovery, account lockouts, and trusted device interactions. The intention was not only to define improved journeys, but also to establish a clear benchmark of what a best-in-class experience should look like.
This project was heavily research-driven, combining multiple data sources to build a comprehensive understanding of user behaviour and pain points.
Quantitative and qualitative inputs included behavioural analytics, customer feedback platforms, and app store reviews, which were synthesised to identify recurring patterns such as failed login attempts, confusion around credential recovery, and lack of clarity in error messaging.
These insights were further validated through UserTesting.com using both moderated and unmoderated approaches. Moderated sessions allowed for deeper probing into user expectations and mental models, while unmoderated tests provided broader behavioural validation across key tasks.
A competitive analysis was also conducted across telecom and adjacent industries to benchmark best practices in authentication flows, particularly around edge cases such as account lockouts, recovery journeys, and trusted device management.
To support the synthesis of large datasets, I cautiously used ChatGPT under my supervision to cluster findings, identify patterns, and accelerate insight generation, ensuring all outputs were manually validated before informing design decisions.
The research was continuously structured into themes, problem statements, and opportunity areas, which directly informed the design direction.
At the same time, SMARTY was evaluating a packaged authentication solution, which introduced constraints around what could realistically be implemented.
The high-fidelity designs therefore acted as a reference framework — a “north star” — helping guide decisions and ensuring that the final implementation remained as close as possible to user needs.
The work provided a structured and validated foundation for improving authentication journeys, aligning user expectations with technical feasibility. It also enabled more informed decision-making when assessing external solutions, ensuring that user experience considerations remained central throughout.
I worked on the evaluation of a new webchat solution that had been selected for implementation, identifying significant usability, accessibility, and content-related issues that impacted its suitability.
Early in the process, I raised concerns regarding the quality of the solution, particularly in areas such as accessibility compliance, content truncation, article formatting, and overall user experience. Initial feedback did not gain immediate traction, requiring a more structured and evidence-based approach to build alignment.
The evaluation was grounded in a structured heuristic and accessibility review of the platform, assessing it against recognised usability principles and expected interaction standards.
This included identifying issues such as inconsistent interaction patterns, lack of clarity in content hierarchy, and accessibility gaps that could impact users relying on assistive technologies. Particular attention was given to how content was presented within the chat experience, where truncation and formatting issues affected readability and comprehension.
To support analysis at scale, I cautiously used ChatGPT under my supervision to help consolidate findings and structure insights, ensuring outputs remained accurate through manual validation.
The findings were iteratively refined and mapped to user and business impact, enabling clearer communication of risks.
As the evaluation progressed, I compiled a comprehensive set of findings that highlighted the impact of these issues on both user experience and operational efficiency.
These were translated into a structured narrative and presented to senior stakeholders to support decision-making.
As a result of this work, approximately 50% of the proposed solution was de-scoped due to not being fit for purpose. In parallel, the vendor began implementing several of the recommendations raised during the evaluation, improving aspects of their product and reducing the level of customisation required internally.
I worked alongside two other designers to define and build a new design system from the ground up.
This initiative was driven by the need to address significant inconsistencies and scalability issues present in the existing ecosystem.
Before starting, it was clear that the organisation had gone through multiple attempts at establishing a design system, with seven previous iterations contributing to inconsistencies across both the system itself and the live product. Different areas of the website reflected different historical versions, creating fragmentation in both design and implementation.
Our approach focused on creating a unified and scalable system that could support current needs while remaining flexible for future evolution. This involved auditing existing patterns, identifying redundancies, and consolidating components into a coherent structure.
A key concept introduced was the use of “recipes” — a flexible layer designed to capture edge cases and more complex compositions. These acted as a sandbox for patterns that were not yet standardised, allowing them to be tested and validated before being formally incorporated into the system if they proved reusable.
The work was informed by a combination of internal audits, developer collaboration, and best practice benchmarking. Existing components in Storybook were reviewed in detail, identifying inconsistencies, duplication, and technical constraints that needed to be addressed.
Close collaboration with engineering teams ensured alignment between design and implementation, particularly when rationalising legacy components and redefining patterns.
To support synthesis and organisation of large volumes of system-related inputs, I cautiously used ChatGPT under my supervision to structure findings and documentation, ensuring outputs were always validated before adoption.
Significant effort was dedicated to establishing strong foundations across typography, colour, spacing, and responsive behaviour. Design tokens and variables were defined to ensure consistency and scalability, while responsive breakpoints were standardised to support both web and mobile contexts.
The system was built using Figma variables, components, and variants, allowing for greater flexibility and maintainability. Modes were introduced to support different contexts, including responsiveness and mobile app patterns, ensuring the system could scale across platforms.
Accessibility considerations were embedded into the system, including colour contrast, interaction states, and component behaviour, aligning with best practices and ensuring consistency across experiences.
Governance played a critical role in ensuring long-term success. Clear documentation was created to support usage, while collaboration with developers helped align Storybook implementation with design standards.
Efforts were made to reduce fragmentation by consolidating legacy components and establishing clearer guidelines for contribution and evolution of the system.
The result was a scalable and cohesive design system that addressed historical inconsistencies while providing a strong foundation for future development. It enabled improved collaboration between design and engineering, reduced duplication, and created a more consistent user experience across the product.