Lowlytics
UX Designer for a comprehensive design of a low-code platform aimed at empowering users to build, monitor, and optimise AI-powered applications. I also contributed to designing a scalable design system and led user research to ensure an intuitive, user-friendly experience.
Lowlytics is a low-code platform that empowers users to build, monitor, and optimize AI-integrated applications — without writing complex code.
Our goal was to bridge the gap between powerful AI tools and users with minimal technical expertise, creating an environment that feels intuitive, reliable, and empowering.This project was created for a client I worked with at Gravitad. Due to an NDA, the case study has been renamed and anonymised.
_edited.png)
_edited.png)
.png)
The Challenge
AI app-building is traditionally a technical, opaque, and overwhelming process — filled with confusing errors, endless documentation, and little guidance for new users. For non-developers and startup founders trying to build AI-powered experiences, this complexity became a major barrier.
How might we simplify the AI app development journey, making it accessible and transparent for users of all skill levels?
My Role
Approach
-
​Designing core user experiences for app building, monitoring, and validation
-
Creating a scalable, Tailwind-based design system
-
Collaborating closely with engineering and product teams
-
Supporting usability testing and iterating based on real user feedback
We grounded our design approach in three pillars:
-
Clarity — Interfaces should anticipate user needs and reduce decision fatigue.
-
Guidance — Provide contextual support without overwhelming users.
-
Trust — Surface AI results transparently, without black-box experiences.
User Research
To ensure we were building the right solution, we conducted in-depth interviews with 10 early adopters from various backgrounds:
-
4 startup founders
-
3 product managers
-
3 junior developers
Throughout these discussions, several recurring pain points became evident. One founder mentioned, “I don’t know where to start.” A product manager shared, “The error messages are too technical for me to understand and fix.” Also, a junior developer expressed, “I’m scared I’ll break something.”
These insights were crucial, highlighting a clear gap in the market for a solution that went beyond mere functionality. Users weren’t just looking for a tool that worked, they needed an intuitive platform that provided clear guidance and support at every step. This research shaped the core focus of the design: to make the process seamless, transparent, and empowering for both technical and non-technical users.
Crafting the user flow
Based on the research, we developed a simplified journey that included five core stages:
-
Creating a project—clear, easy-to-follow steps for setting up an app.
-
Adding AI modules — A drag-and-drop interface for quick integration.
-
Monitoring build progress — Real-time updates on the status of the project.
-
Validating results — Providing users with actionable feedback and next steps.
-
Optimising and relaunching — Tools to improve and enhance app performance.
.png)
.png)
.png)
.png)
Design System
To ensure consistency and scalability across the platform, I created a lightweight design system using Tailwind CSS. It included reusable components, defined spacing rules, colour tokens, and typography styles that aligned with the overall tone of the product—modern, functional, and accessible. Building the system early on helped speed up screen design and made it easier for developers to implement UI elements with clarity and precision.
Turning documentation into a meaningful experience
The original documentation page surfaced raw logs and vague system messages, which often left users—especially non-technical ones—uncertain about what needed attention.
​
One of the most critical areas was Documentation & Debugging, users received raw JSON logs and vague error reports — unusable for anyone without a strong tech background.
I redesigned this experience around two key elements:
-
Validation Checklist — A prioritized, human-readable list of errors and warnings
-
AI Debugging Insights — Simple, actionable recommendations (e.g., “Add an aria-label to your primary button”)
These changes helped users feel informed, in control, and capable of fixing problems on their own.
UI Design
_edited.png)
.png)
Testing
As part of our iterative design process for Lowlytics, we conducted usability testing with 10 early adopters, including startup founders, product managers, and junior developers. We chose a mix of moderated remote sessions and unmoderated self-guided tests to capture a wide range of feedback.
​
Key Insights:​​​
-
Pain Point: Error messages were too technical and caused frustration, especially during the AI module configuration.
-
Change: We rephrased error messages in non-technical language, offering clear action steps instead of jargon, making them more accessible.​​​
-
Pain Point: Users struggled with understanding the results of their AI models during testing.
-
Change: Introduced colour-coded feedback and simplified success/failure messages, along with quick actions like "Retry" and "Adjust parameters."
Conclusion
After implementing the design changes based on our user testing insights, we began to see tangible results. For example, onboarding time dropped significantly—new users were able to complete the setup in under 3 minutes, compared to the previous 4.5 minutes. This change was a direct result of simplifying the flow and providing clearer guidance right from the start.
​
In addition, user engagement improved by 15%, with more users successfully completing the app creation process. By focusing on intuitive design and simplifying error messages, we made it easier for users to stay on track and finish what they started.
​
We also saw a 40% increase in the number of users who successfully ran tests and optimised their models. The clearer, colour-coded feedback helped users take immediate action, building their confidence in the platform.
​
The design improvements didn’t just benefit the users; they also had a positive impact on our team. With clearer workflows and a more intuitive interface, the team saved around 5 hours per week in app-building and deployment time, leading to greater efficiency overall.
Ultimately, these changes reinforced the value of a user-centred design approach, making Lowlytics not only more user-friendly but also more effective and efficient, especially for non-technical users.