Evorayn

Inspiration

Trading can be overwhelming. Traders move between charts, news, indicators, broker apps, and AI tools, trying to figure out what actually matters.

We wanted to build something that brings those pieces together and helps traders make more informed decisions without blindly handing control to AI.

That idea became Evorayn.

What it does

Evorayn is an AI-powered, non-custodial trading intelligence platform.

It combines market research, quantitative analysis, historical performance data, and risk controls to help traders understand opportunities and make better decisions.

Instead of giving users vague AI predictions, Evorayn is designed to support its analysis with actual data and historical outcomes.

Users stay in control of their funds through their existing brokers and exchanges. Evorayn can support research-only insights, AI-assisted trading, and eventually controlled automation with strict risk limits.

How we built it

Evorayn is built with:

• Flutter for the mobile app
• Python and FastAPI for the backend
• PostgreSQL for persistent data
• Redis for caching and temporary state
• AI models for market research and analysis
• A broker-adapter architecture for connecting brokers and exchanges

One of the most important parts of the architecture is keeping AI analysis separate from trade execution.

AI can research and suggest opportunities, while deterministic systems handle things like authorization, risk limits, validation, and execution safety.

Challenges we ran into

One of the biggest challenges was figuring out how much control AI should have.

It is relatively easy to ask an AI to analyse a market and suggest a trade. It is much harder to build a system where AI can be useful without being allowed to bypass risk controls or make unsupported claims.

We also had to think carefully about authentication, session security, unreliable networks, broker connectivity, and what should happen when the system cannot confirm whether an order was successfully submitted.

In trading, simply retrying an uncertain request could result in a duplicate trade, so failure handling has to be designed very carefully.

Accomplishments that we're proud of

We are proud that Evorayn is being designed around evidence and risk management instead of AI hype.

AI-generated insights are intended to be supported by real market data, historical statistics, and clearly defined risk controls.

We are also proud of the non-custodial approach. Evorayn does not need to hold a user's money. Users keep their funds with their broker or exchange.

Another accomplishment has been building the foundation as a real product, including the mobile app, backend, authentication, onboarding, security rules, and architecture needed for the trading features that come next.

What we learned

The biggest thing we learned is that building an AI trading platform is not mainly an AI problem.

The difficult parts are data quality, risk management, security, reliability, execution safety, and making sure the system behaves correctly when something goes wrong.

We also learned that AI should not pretend to know things it cannot know. Instead of inventing confidence scores, Evorayn is designed to show measurable evidence such as historical outcomes wherever possible.

What's next for Evorayn

Next we want to expand Evorayn into the full trading experience.

That includes live market data, interactive charts, paper trading, broker connections, portfolio tracking, quantitative analysis, AI-powered market research, strategies, alerts, and eventually carefully controlled automated trading.

The goal is to build a platform where AI helps traders make better decisions without replacing discipline, evidence, or risk management.

Built With

• dart
• fastapi
• flutter
• postgresql
• typescript