TriadFi

Problem Statement

We have built an AI-driven trading platform designed for institutional adoption, where investors can allocate capital and trade with the assistance of autonomous AI agents.https://github.com/Triad-Finance/triad-fiAt the heart of the platform are three specialized agents that work together to identify opportunities, manage risk, and execute trades seamlessly on decentralized exchanges such as 1inch:🔹 AgentsSignal Agent – Continuously analyzes market data, applying AI-driven technical indicators and strategies (e.g., RSI, MACD, trend models) to generate actionable trade signals.Risk Agent – Validates each proposed trade against predefined risk frameworks (exposure caps, volatility filters, portfolio allocation rules) to ensure compliance and safety.Execution Agent – Determines the optimal time and limit price for each order. It then places and manages the trade directly on 1inch DEX, ensuring best execution with minimal slippage.🔹 Asset Managers (AMC)Beyond investors, the platform enables AI-assisted Asset Managers (AMCs) to design, deploy, and manage trading strategies. These managers can:Encode strategies into the agent framework.Leverage AI reasoning (e.g., symbolic + neural methods) to adapt to market conditions.Offer whitelabeled strategies to institutions, enabling tailored products for hedge funds, family offices, and DAOs.🔹 Key FeaturesAI-Driven Trading → Smart market analysis with autonomous decision-making.Risk-First Design → Built-in constraints and compliance checks for institutional trust.Limit Order Execution → Dynamic placement on 1inch with adaptive pricing.Whitelabel Ready → Customizable for financial institutions and asset managers.Dual Ecosystem → Investors provide capital; AMCs provide strategies, creating a marketplace of AI-driven funds.

Solution

Our platform combines AI reasoning, autonomous agents, and DeFi protocols to create an institutional-grade trading framework.🔹 Core ArchitectureFetch.ai uAgents → The backbone of our system, enabling lightweight, autonomous, and message-driven AI agents.Three-Agent Workflow:Signal Agent – Built with Python and integrated technical indicators (RSI, MACD, ML models) to generate trade signals based on Swap Token data from TheGraph Token APIRisk Agent – Encodes institutional risk constraints (exposure, volatility, liquidity) using both rule-based logic and symbolic reasoning (via MeTTa – Meta Type Talk). Data based on TheGraph Token API and CoinDeskExecution Agent – Uses the 1inch Limit Order Protocol to place buy/sell orders at optimal prices, managing order lifecycle and slippage protection.🔹 Strategy Layer (AMCs)Asset Managers (AMCs) can design strategies directly in MeTTa or Python, which plug into the Signal + Risk pipeline.Strategies are modular and whitelabel-ready for institutions to adopt or customize.🔹 AI Reasoning & Risk ManagementMeTTa (Hyperon’s reasoning language) is embedded inside agents to represent constraints and adapt strategies dynamically.Combines neural (ML) + symbolic (MeTTa) AI for explainable decision-making with reliable data from TheGraph's Token API🔹 DeFi & On-Chain Execution1inch DEX Integration → Secure limit order placement via smart contracts.Wallet Integration → Each agent signs and submits transactions using EVM-compatible wallets.Transparency → Attested trade logs on-chain for investor and institutional auditability.🔹 Whitelabel ProductModular architecture makes it easy to rebrand, configure, and deploy for hedge funds, DAOs, and family offices.Institutions can plug in their own strategies while relying on our agent framework + execution layer.✅ Tech Stack HighlightsAgents: Fetch.ai uAgents (Python SDK)Reasoning: MeTTa (OpenCog Hyperon) for symbolic AITrading: 1inch Limit Order Protocol, EVM walletsData: On-chain (TheGraph) + off-chain (Coindesk) price feeds, technical indicators and historical social dataDeployment: Whitelabel SaaS/DeFi hybrid for institutions