The problem AetherGuard solves

AetherGuard — Privacy-First AI Portfolio Guardian

Overview

AetherGuard is a platform that helps users get AI-powered trading suggestions while keeping their investment strategies completely private.

Instead of sending strategies to AI providers, AetherGuard only shares the user’s investment goal. All actual strategies and guardrails remain private, yet cryptographically verifiable.

The system leverages:

• Merkle proofs to verify strategy compliance
• Fileverse encrypted documents for secure and decentralized storage

This ensures AI can assist users without ever seeing their private strategy.

---

The Problem

Most AI trading assistants today require users to send:

• Portfolio details
• Investment strategies
• Risk and trading rules

to centralized systems, which creates major risks:

Even if users want privacy, traditional tools do not allow strategy confidentiality.

---

Real-World Example

A trader manages $100,000 crypto portfolio with these rules:

Without AetherGuard: AI platforms require sending the rules → full strategy is visible.

With AetherGuard: AI only sees the goal and portfolio state. It can propose trades but cannot access private rules.

Example workflow:

---

The AetherGuard Solution

1. Strategy Creation

Users define:

• Investment Goal → Sent to AI
• Private Guardrail Rules → Stored in Fileverse encrypted documents

Rules are converted into a Merkle tree, and the root hash is used for verification.

---

2. AI Market Suggestions

The AI only receives:

• Current portfolio state
• Investment goal

It analyzes market data and generates suggestions:

• Example: “Buy 20% ETH with available cash”

AI never sees strategy rules.

---

3. Guardrail Verification

Before execution:

• Each AI suggestion is checked against the Merkle root
• Violations trigger UI warnings
• Users can approve or reject trades on-chain

---

4. Permanent Audit Trail

All strategy and decision data are stored in Fileverse:

This ensures transparent, tamper-proof auditability.

---

5. Collaborative Multi-Sig

AetherGuard enables team portfolio management:

• Teams create a shared Fileverse document
• AI proposals are reviewed by multiple members
• Execution can require multi-signature approval

This is ideal for DAOs, treasuries, or collaborative trading teams.

---

Why AetherGuard

AetherGuard allows users to:

• Maintain private strategies
• Receive AI market insights
• Verify AI suggestions cryptographically
• Approve trades on-chain
• Keep a permanent decentralized audit trail

It combines AI assistance with full privacy, security, and user control, making AI-powered trading safe and trustworthy.

Challenges we ran into

Challenges We Faced

Building AetherGuard involved solving several key challenges around privacy, collaboration, and portfolio simulation.

---

1. Verifying AI Suggestions Without Revealing Strategy

We wanted the AI to suggest trades without ever seeing the user’s private strategy.
Normally, AI systems require the full strategy to know if a trade is allowed, but sending it to the AI would break privacy.

Solution:
We used Merkle Root Proofs:

• Each strategy rule is hashed and stored in a Merkle tree.
• The Merkle root acts as a fingerprint of the strategy.
• AI suggestions are checked against this root to verify compliance without revealing the rules.
• We also show the verification process in the UI for transparency.

---

2. Collaborative Multi-Signature Trading

For team-managed portfolios or DAOs, multiple people need to review and approve trades.

Challenge:
How to enable secure team collaboration while preserving privacy and avoiding a single point of failure.

Solution:
We built Fileverse-based multi-sig boardrooms:

• A team document stores proposals.
• Members can review suggestions.
• Trades require multi-signature approval before execution.

This ensures no single person—or AI—can move funds independently.

---

3. Portfolio Simulation & Visualization

As students, we don’t have access to real mainnet funds or large testnet portfolios.

Challenge:
How to demonstrate trading flow and AI decision-making realistically without risking real assets.

Solution:

• We use a simulated portfolio, starting with 100% cash.
• When AI suggests a trade (e.g., buy 20% ETH), the portfolio updates virtually: 20% ETH, 80% cash.
• AI can read this simulated portfolio to propose strategically meaningful trades.
• Execution still requires user approval, ensuring safety.

This allows us to show the full AI workflow in the demo even without real funds.

---

Summary

The three main challenges we overcame:

1. Strategy privacy vs verification — solved with Merkle proofs.
2. Collaborative multi-sig trading — solved with Fileverse team documents.
3. Portfolio simulation for demo purposes — solved with a safe cash-to-simulated-asset system.

These solutions ensure that AetherGuard maintains privacy, security, and transparency, while still providing a realistic AI-assisted trading experience.