Inspiration

TradeWalls addresses the need for a trading strategy that balances active and passive management, offering a structured "planned trading" approach to help users accumulate assets while minimizing emotional decision-making.

What it does

TradeWalls enables users to plan trades by defining "walls" for cryptocurrency pairs. An autonomous agent monitors the market and alerts users when trades are possible based on these rules. AI suggests trading walls based on user input.

How we built it

TradeWalls was built using NEAR BOS and deployed to the NEAR AI.Nexus. A self-hosted open-source app polls prices, stores trades, and alerts users when trade walls are triggered. The backend uses Flask, and the frontend is built with React.

Challenges we ran into

Automating trades across different exchanges is complex due to varying prices and potential edge cases. Advanced agentic AI could address this in the future. We settled on trade alerts instead of automated trading.

Accomplishments that we're proud of

We refactored our existing trade strategy code, which we have used for years, into this hackathon entry. We built and open-sourced the backend and integrated AI suggestions for trade walls by parsing JSON responses.

What we learned

Developing TradeWalls provided insights into the challenges of creating an autonomous cryptocurrency trading tool. We also explored how AI-assisted suggestions can enhance human-in-the-loop rule-based autonomous agents.

What's next for TradeWalls

We will continue using TradeWalls internally to structure our trades. Future plans include exploring AI-based trade optimization and pathfinding across multiple exchanges to enhance the platform's capabilities.

Built With

Share this project:

Updates

posted an update

The bounties this is applying to are: NEAR AI Nexus - listed on the Nexus with LLM interaction Gensys - llms create an interface that enables planning for the future of trades Marlin Bounty 2 (not sure if eligible)

Log in or sign up for Devpost to join the conversation.