Skip to content

hgunduzoglu/neurotrade.eth

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

42 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

neurotrade.eth

AI-assisted cross-chain trading interface that combines conversational market analysis with explicit, user-signed execution through 1inch Fusion+.

1inch Partner Prize · ETHGlobal Cannes 2025

The assistant analyzes and recommends. The user chooses the trade and signs the order.

Overview

neurotrade.eth was built to make token research and cross-chain execution part of one coherent flow.

Users can ask the assistant about supported assets, review technical signals and risk indicators, inspect their portfolio, and continue into a cross-chain swap without leaving the application.

Analysis and execution are deliberately separated. The assistant cannot move funds or execute trades autonomously. Every swap is configured and signed explicitly by the user.

The project was built by a team of two for ETHGlobal Cannes 2025 and received a 1inch partner prize.

Core Features

  • Conversational token and market analysis
  • Buy, sell, and hold signals with risk context
  • Cross-chain swaps through 1inch Fusion+
  • Privy email and wallet authentication
  • Embedded wallets for users without an existing wallet
  • External wallet support through wagmi
  • Multi-chain portfolio and transaction views
  • Server-side proxying for protected API credentials
  • Explicit user approval and signing before execution

How It Works

User
  |
  v
Next.js Application
  |
  |-- Chat and token analysis ------> FastAPI AI service
  |                                    |
  |                                    `--> Market data and indicators
  |
  |-- Portfolio data ---------------> The Graph
  |
  |-- Wallet onboarding ------------> Privy + wagmi
  |
  `-- Cross-chain swap -------------> Server-side API proxy
                                       |
                                       `--> 1inch Fusion+

Analysis Flow

Natural-language request
        |
        v
Token and market data
        |
        v
Technical indicators
RSI · SMA · momentum · volatility · volume
        |
        v
Signal and risk assessment
BUY · SELL · HOLD
        |
        v
Explanation presented to the user

Swap Flow

Select source and destination assets
        |
        v
Request Fusion+ quote
        |
        v
Build intent-based order
        |
        v
Sign EIP-712 order
        |
        v
Submit to the 1inch relayer
        |
        v
Poll order status
        |
        v
Reveal the matching secret when required
        |
        v
Cross-chain settlement

My Contribution

I owned the third-party integration layer that connected the product experience to its external infrastructure.

My work included:

  • Integrating the 1inch Fusion+ quote, build, signing, submission, status, and secret-reveal lifecycle
  • Building server-side proxy routes so 1inch credentials were never exposed to the browser
  • Connecting Privy authentication and embedded wallet onboarding
  • Supporting external wallets through wagmi
  • Integrating portfolio and transaction data through The Graph
  • Connecting market-data feeds used by the interface and analysis service
  • Wiring the frontend to the Python AI service
  • Normalizing external API responses and failure states for the application

Key Technical Decisions

Analysis and execution are separate

The assistant produces signals, explanations, and risk assessments, but it cannot execute a transaction.

A trade only proceeds after the user opens the swap interface, chooses the parameters, and signs the order. This boundary is enforced by the product flow rather than being left as a disclaimer.

Fusion+ instead of a basic swap request

1inch Fusion+ uses signed, intent-based orders and resolver-driven execution. The integration supports cross-chain settlement while reducing the transaction's exposure to common forms of MEV.

The resulting flow is more involved than a conventional swap API: the application must construct and sign an order, track its status, and reveal the appropriate secret during execution.

Provider credentials remain server-side

Authenticated 1inch requests pass through Next.js API routes. The browser communicates only with application-controlled endpoints and never receives the provider credential.

This also creates a single layer for:

  • Error normalization
  • Request validation
  • Provider changes
  • Rate-limit handling
  • Logging and observability

Embedded wallets reduce onboarding friction

Privy allows a user to create an embedded wallet through email authentication while still supporting users who prefer MetaMask, WalletConnect, or Coinbase Wallet.

This allows new users to move from login to a signed order without first installing a browser extension.

Technology Stack

Area Technologies
Frontend Next.js 14, React 18, TypeScript
Server state TanStack Query
Authentication Privy
Wallet integration wagmi, viem, ethers.js
Cross-chain execution 1inch Fusion+
AI service Python, FastAPI
Market analysis RSI, SMA, momentum, volatility, volume
Portfolio data The Graph
Data visualization Chart.js
API protection Next.js API routes
Optional deployment proxy Cloudflare Workers

Running Locally

Prerequisites

  • Node.js
  • npm
  • Python 3
  • A Privy application
  • A 1inch developer API key
  • Optional Graph API credentials

1. Clone the repository

git clone https://github.com/hgunduzoglu/neurotrade.eth.git
cd neurotrade.eth

2. Configure the environment

Copy the example environment file:

cp .env.example .env

Configure the required values:

NEXT_PUBLIC_PRIVY_APP_ID=your_privy_app_id
NEXT_PUBLIC_AI_API_URL=http://localhost:8000

ONEINCH_API_KEY=your_1inch_api_key
GRAPH_API_KEY=your_graph_api_key

AGENT_SEED=neurotrade_ai_agent
AGENT_PORT=8000
LOG_LEVEL=INFO
USE_AGENTVERSE=true

Do not expose ONEINCH_API_KEY through a NEXT_PUBLIC_ environment variable. It is read only by the server-side proxy.

3. Start the application

The repository includes a launcher that installs the frontend and Python dependencies and starts both services:

python start_neurotrade.py

The services will be available at:

Frontend: http://localhost:3000
AI API:   http://localhost:8000

Manual Startup

Install the frontend dependencies:

npm install

Install the AI service dependencies:

python -m pip install -r neurotrade_ai_agent/requirements.txt

Start the AI API:

python -m uvicorn neurotrade_ai_agent.api_bridge:app \
  --host 0.0.0.0 \
  --port 8000 \
  --reload

In another terminal, start the frontend:

npm run dev

Project Context

  • Event: ETHGlobal Cannes 2025
  • Team size: Two
  • Result: 1inch partner prize
  • My role: Third-party API and wallet integrations
  • Status: Hackathon prototype

Disclaimer

neurotrade.eth is an experimental hackathon project. Its market signals are informational and should not be interpreted as financial advice. Do not use the prototype with funds you cannot afford to lose.

About

AI-assisted cross-chain trading interface with explainable market signals, Privy wallets, and MEV-protected swaps through 1inch Fusion+ — 1inch prize winner at ETHGlobal Cannes 2025.

Resources

Stars

0 stars

Watchers

0 watching

Forks

Releases

No releases published

Packages

 
 
 

Contributors