My Role & Responsibilities
In this MVP stage, my responsibilities encompassed UX/UI design, research, and stakeholder collaboration. I planned user journeys, built wireframes and interactive prototypes, and worked closely with developers, data scientists, and a project manager to align on technical feasibility. Daily stand-ups and weekly workshops kept our efforts agile and user-focused, while my attention to data clarity and accessibility shaped the visual system into something both intuitive and scalable.
The Problem
Many organisations struggle to make fully informed decisions due to outdated or disconnected data. eamli’s AI technology aimed to resolve this by predicting diverse outcomes and offering preemptive insights. However, the existing proof of concept needed a streamlined, comprehensible user experience to resonate with high-level stakeholders in enterprise and government contexts.
Without robust UX principles, even the most advanced AI model can leave users confused. My priority, therefore, was to design an interface that made these predictive insights instantly usable, as well as easily comparable across multiple hypothetical scenarios.
Why It Matters
Government agencies and large enterprises frequently deal with decisions involving multi-million-pound contracts and decade-long implications. For them, an outdated or confusing tool can translate into missed opportunities and significant financial risk. eamli’s mission was to reduce guesswork through real-time data modelling, so ensuring that the platform was accessible and informative at first glance became critical.
Objectives & Discovery
The primary objective was to create a tangible MVP that balanced AI sophistication with user-friendliness. We also sought to validate our approach by collecting feedback from “friendly” user groups, typically partner organisations that were open to testing the product.
During initial discovery, I helped to conduct a series of informal interviews with internal teams and these partner organisations. Early prototypes and user journeys were shared to understand the pain points and immediate desires: a need for clear, concise data outputs and simple workflows for scenario creation. Competitive analysis further revealed that most existing solutions relied on manual modelling, so fully automating and visualising potential outcomes was a clear differentiator for eamli.
Product Architecture & Planning
Over a four-month period, we iterated on the MVP, refining its core features:
- Scenario Building & Comparison – Users needed to define various business models quickly, adjusting inputs without navigating convoluted menus.
- Data Visualisation & Reporting – Clear charts, graphs, and scorecards helped stakeholders see at a glance where each scenario might lead.
- Organisation Management – Teams wanted the ability to manage multiple decision-makers and datasets under one unified platform.
Throughout this phase, I maintained close communication with data scientists. Since the back-end logic was highly flexible—capable of handling numerous company structures—I focused on designing a modular UI that could adapt to evolving or custom data inputs without overwhelming the user.
Wireframing & Prototyping
Using Figma for wireframing and interactive prototypes, I shared multiple iterations with both internal and external stakeholders. Each testing round combined task-oriented evaluations (e.g., “create a new scenario and compare it to an existing one”) with broader feedback discussions about navigation and overall clarity.
These sessions confirmed that our approach to minimalist data visualisation was effective. However, participants requested even faster ways to compare outcomes and toggling between different hypothetical scenarios. It was at this point we recognised that future versions might benefit from natural language modelling, which the eamli team later pursued after I transitioned off the project.
Visual Design & Accessibility
While we followed foundational brand guidelines, most UI components were custom-designed to handle the platform’s data complexity.
To maintain a clean, modern feel, each dashboard view minimised clutter and featured easy-to-read charts. I embedded interactive tooltips and guided modals where necessary, helping users discover advanced features without feeling overwhelmed.
Ensuring AA-level accessibility was essential for text contrast, colour usage, and consistent patterns. Given the robust nature of scenario comparisons—often involving various graphs, metrics, and calculations—we focused on a desktop-first experience, though we explored mobile-responsive layouts for stakeholders who needed quick updates on the go.
Key Challenges & Lessons
One notable challenge was translating complex AI modeling into straightforward, visual summaries. This required deep collaboration with data scientists to ensure that the system’s intelligence was neither lost in simplistic displays nor buried in technical jargon.
For me, this project was also a first major exploration into data visualisation, confirming how iterative testing is vital to confirm design efficacy. Incorporating user education into the UX—through training prompts, tooltips, and contextual tips—proved pivotal in demystifying the AI output for less technical decision-makers.
Outcome & Impact
By the end of this 2019–2020 phase, we had established a functional MVP that allowed businesses and government stakeholders to experiment with various “what-if” scenarios. This work provided the foundations for subsequent iterations of eamli, which went on to adopt features like natural language modeling and more advanced analytics. While I wasn’t directly involved in those later phases, the MVP design contributed to notable milestones, including a collaboration with IBM and adoption by a UK Government Ministry for long-term procurement planning.
Eventually, eamli’s continued evolution won Digital Innovation of the Year at the Digital DNA awards, validating how a strong design approach can supercharge cutting-edge AI solutions in the marketplace.
Personal Highlights & Mentorship
One aspect I’m especially proud of is how this MVP transformed a high-level concept into a practical tool. By focusing on user workflows and data clarity, we proved the viability of AI-assisted scenario building—instilling trust that an algorithm could, in fact, simplify day-to-day strategic decisions.
I also had the opportunity to onboard and mentor a junior designer during this period, guiding them through Figma best practices, lean workshops, and UX fundamentals. They took on increasing responsibilities as the product developed, eventually becoming a core contributor to future eamli versions.
Tools & Techniques
- Figma: Primary wireframing and prototyping platform, where I managed a component library for consistent UI elements.
- Miro: Used for collaborative brainstorming, journey mapping, and stakeholder workshops.
- Lean UX: Employed short feedback loops and agile ceremonies, ensuring continuous refinement and stakeholder alignment.
Looking Ahead
Designing this 2019–2020 version of eamli reinforced my belief that even the most advanced AI technology must remain user-centred to achieve real-world impact. By prioritising modular interfaces, clear data insights, and ongoing user feedback, we created a platform that both recognised and adapted to the varying needs of enterprise and public-sector customers.
I continue to draw on these lessons in my subsequent roles, especially when bridging the gap between complex back-end systems and the people who rely on them for strategic, future-focused decisions. If you’d like more details on this process or how I can bring similar value to your team, feel free to reach out anytime.
