Decentralized AI and Blockchain
Defined Macrocosmos’s design system and AI interfaces, transforming technical complexity into simple, human-centered experiences.
What is Macrocosmos
Macrocosmos.AI is an open-source AI ecosystem that enables developers to build, train, and deploy intelligent systems on decentralized infrastructure. It connects data, tools, and models through specialized subnets, empowering a global community to create scalable and transparent AI solutions.
Scope of work
Defined the visual and interaction framework for Macrocosmos’s AI products, supported by extensive research, competitive analysis, and created pricing strategy. This document highlights one of the key features designed as part of that process.
Feature: Data Collection & Build Flow
In Macrocosmos, users collect data from platforms like Reddit, Youtube, or X to train AI models and analyze trends. Data collection gathers raw information, while data building structures it into usable datasets ready for insight or deployment.
When I joined the project, the existing Data Collection interface was already developed by the engineering team. Our goal was to add more functionality and scalability while improving clarity for users making data-driven decisions. The original version offered limited information, making it difficult for users to decide whether to build a dataset or not. Below is the image of what was already there
Through collaborative discussions with developers and users, I realized that the issue went beyond visuals it was about information hierarchy and user flow. I began by mapping the user persona, entire experience and exploring multiple flow options to better structure the process.
The first redesign attempt consolidated too much information on a single screen, which made the interface cluttered and confusing. Also gave me opportunity to guide the team with design system to bring consistent visual design language.
Design System
I created a living visual language and design system for Macrocosmos tailored to a tech-savvy audience, primarily developers. The design aimed to convey a futuristic vision aligned with the brand’s AI-driven identity. Inspired by the concept of the cosmos reflected in the name the system adopted a dark theme, resonating with developers.
Final Design
We adopted a multi-step flow approach separating key actions into clear stages. Users could now: 1. View a task library with all their ongoing or completed collections. 2. Dive into task details to view progress, insights, and costs. 3. Build datasets confidently, supported by transparent pricing and status visibility. The result was a clean, scalable interface that balanced complexity with usability giving users full control over their data collection process while maintaining a futuristic, developer-focused design language.
Challenges
Early in the process, it was unclear what information users actually relied on when deciding to build a dataset. The interface needed to display multiple layers of data cost, status, platform, and record size without overwhelming the user. Finding the right balance between information depth and visual clarity was a constant iteration process. Introducing pricing to a technical workflow was tricky it needed to be clear, trustworthy, and non-intrusive. Users had to understand cost implications instantly without feeling interrupted or confused by the interface.
Impact
Our initial goal was to generate $10K in revenue within the first week, marking Macrocosmos’s first monetization milestone. With a clear value proposition and an intuitive design that helped users understand costs and actions instantly, we reached that target in just 2 days. Encouraged by this success, the team set a new goal $100K in the next two weeks positioning design as a key driver of growth.
