Replicats is a platform where users can hire AI agents to manage crypto portfolios autonomously. The system allows users to add funds, adjust risk profiles (from conservative to aggressive), and track market predictions based on news and data. The agent operates 24/7, protecting and growing returns.
Abstract
The crypto investment market has seen fast growth in the use of autonomous agents. However, many solutions still deliver confusing interfaces, low predictability, and little transparency about the agent’s decisions. This represents a clear business opportunity: offering a platform where users trust and at least understand what their agent is doing.
Speed of execution was a key point in this project. We needed an interface that made activating the agent quick and intuitive — mainly through chat, which is the primary channel for interacting with the AI. In addition, the platform needed to support multiple agents, allow general account and fund settings, and give users clear and direct control over their investment profile.
With this scenario in mind, I structured the design process prioritizing information clarity, fast main flows, and a technical foundation that would allow continuous product evolution.
Process
Understanding the challenge
The main challenge was to deliver a functional interface on a short timeline, while also needing to learn about the crypto market and automated investments. We also had a clear technical rule: use Tailwind and Shadcn as the component base.
Desk research and benchmark
The first step was dedicated to market research and competitor analysis. We looked for references on similar platforms — both in crypto and traditional investment systems — to understand interface patterns, expected flows, and opportunities.
Wireframe validation with the team
The first wireframes were presented and validated internally with the product and development teams. This step was essential to align expectations, adjust flows, and ensure navigation made sense before moving to the visual part.
Style definition and refinement with the dev team
With wireframes approved, we defined the visual styles using Tailwind and Shadcn, ensuring consistency with technical rules. Refinement was done together with the development team to make sure components were feasible and reusable.
Phased project delivery
Due to the short timeline and the evolving nature of the product, we chose phased deliveries. First, the main flows and base structure. In later stages, secondary screens, adjustments, and improvements as the product matured. Each delivery allowed immediate execution by the front-end team.
General follow-up
We maintained close follow-up during development, reviewing implementations and clarifying questions. The front-end team was able to evolve the interface autonomously most of the time, calling us only for unexpected or more complex cases.
Solution
As a result, we delivered a scalable interface that was simple to implement and self-driven by the front-end team. This autonomy provided cost savings and agility for the project. The system is still in progress, with some phases already implemented and others waiting for continuous product evolution.
View of the deposit flow with validation actions, general screens, dark/light mode, and expand/collapse.
