About
DecenterAI is a groundbreaking platform that breaks down the barriers to AI development by making it possible for anyone to train, fine-tune, and deploy their own AI models without deep technical expertise. As the product designer, I focused on turning complex machine learning workflows into clear, guided experiences that empower users to bring their ideas to life confidently.
CHALLENGE
While AI holds incredible potential, its development often feels intimidating and exclusive, reserved for those with strong technical or programming backgrounds. The main challenge was designing an interface that makes highly technical processes like dataset uploading, model training, and fine-tuning not only understandable but approachable, engaging, and confidence-building for first-time AI creators.
Design Approach
As the product designer, I led the user experience strategy and interface design for the platform. My approach began with understanding what users find most overwhelming about AI workflows. Through user interviews and competitive research, I uncovered that creators wanted clarity at every step, visual cues that guide decision-making, and reassurance that they are building correctly.
To address these needs, I mapped out user journeys that guide them seamlessly from idea to deployment. The design incorporated onboarding tooltips, progress trackers, and intuitive forms to break down technical jargon into simple, actionable instructions. I also focused on creating a design language that feels professional yet friendly, avoiding sterile or overly technical aesthetics to keep users engaged and motivated.
The Solution
The homepage was crafted to communicate DecenterAI’s mission clearly while inviting users to begin their AI creation journey with confidence. For the AI creation flow, I designed guided pathways that allow users to choose between uploading custom datasets, using predefined templates, or fine-tuning existing models. Each pathway includes clear explanations, visual previews, and step-by-step progress indicators to reduce cognitive load and prevent user drop-off.
The dataset upload interface was designed to be as intuitive as possible, using drag-and-drop functionality with clear validation feedback. For model training and fine-tuning, I designed status visualisations that communicate complex backend processes in human-friendly terms, giving users a sense of control and understanding even when dealing with abstract machine learning concepts.
Finally, the deployment screens were structured to celebrate user achievements, providing clear next steps and opportunities to integrate their models into real-world applications confidently.
Reflection & Learnings
Working on DecenterAI reinforced my belief that design can unlock powerful technologies for broader audiences. Simplifying AI development into approachable, guided experiences requires not just good UI but deep empathy for what users fear, value, and need to feel confident. This project taught me the importance of combining clarity, encouragement, and visual storytelling to make complex processes feel natural and achievable.
