firstName

lastName

email

message

updatesConsent

Want to join the project hub? Tell me more...

PortalAi

useCase

Join the Waitlist

Deploy AI agents to Telegram and Twitter with wallet integration and API access.

# AI Agent Platform

## What It Is

The AI Agent Platform lets users deploy conversational AI agents to Telegram and Twitter. Agents can respond to messages, execute commands, connect to external APIs, and sign blockchain transactions through an integrated wallet.

Agents maintain conversation context, follow configurable behaviour rules, and can be customised for specific use cases.

---

## The Problem

Running an AI assistant across multiple platforms requires:
- Separate bot infrastructure for each platform
- Manual integration with APIs and data sources
- No native support for blockchain transactions
- Repetitive setup for common patterns

---

## The Solution

The platform provides:
- Single deployment target for Telegram and Twitter
- Pre-built integrations with trading APIs (0x), data providers (Covalent, The Graph), and LLMs (OpenAI, Claude)
- Wallet connection for on-chain actions
- Configurable agent behaviour and response patterns

---

## How It Works

1. Create an agent with name and behaviour configuration
2. Connect platform credentials (Telegram bot token, Twitter API keys)
3. Add API integrations as needed
4. Connect a wallet for blockchain actions
5. Deploy

The agent runs continuously, responding to messages and executing configured actions.

---

## Capabilities

**Platform Integration**
- Telegram: command handlers, group chat support, inline queries
- Twitter: reply bots, quote tweets, DM handling

**API Access**
- 0x Protocol for token swaps
- Covalent for wallet and token data
- The Graph for protocol queries
- OpenAI and Claude for language processing

**Blockchain Actions**
- Transaction signing via connected wallet
- Token swaps and transfers
- Contract interactions
- DAO voting

**Agent Configuration**
- Response tone and style
- Trigger conditions
- Rate limiting
- Context memory settings

---

## Development Status

| Milestone | Target | Status |
|-----------|--------|--------|
| MVP (Twitter/Telegram) | Aug 2025 | Live |
| Wallet-connected actions | Sep 2025 | In Progress |
| Agent template marketplace | Oct 2025 | Planned |
| Token-gated deployments | Nov 2025 | Planned |

---

## Use Cases

**Trading Bots**
Telegram bot that monitors prices and executes swaps via 0x based on configured rules.

**Community Support**
Group chat bot that answers questions using project documentation as context.

**Social Automation**
Twitter bot that responds to mentions with relevant content.

**Notifications**
Agent that monitors on-chain events and sends alerts to Telegram.

AI AGENT PLATFORM

AI integration consulting for businesses. Assessment, implementation, and support.


# AI Consulting

## What It Is

Consulting services for AI integration in business operations. We assess existing systems, design AI solutions, implement them, and provide ongoing support.

---

## Services

**Assessment**
- Business process analysis
- Technical infrastructure review
- Data quality evaluation
- Implementation prioritization

**Solution Design**
- Architecture development
- Integration planning
- Prototype development
- Technical specifications

**Implementation**
- Project management
- Custom development
- Data pipeline construction
- System integration
- Testing and deployment

**Support**
- Performance monitoring
- Model retraining
- Technical support
- Scaling assistance

---

## Industry Experience

| Industry | Applications |
|----------|--------------|
| Financial Services | Risk assessment, fraud detection, credit scoring, compliance automation |
| Healthcare | Patient flow optimization, image analysis, treatment recommendations |
| Retail | Demand forecasting, inventory optimization, pricing, recommendations |
| Manufacturing | Predictive maintenance, quality control, production scheduling |
| Professional Services | Document processing, knowledge management, resource allocation |

---

## Engagement Models

**Advisory**
- Strategy sessions
- Technology assessment
- Roadmap development

**Project-Based**
- Fixed-scope implementations
- Proof-of-concept development
- Architecture design

**Retained**
- Ongoing advisory
- Continuous optimization
- On-call support

---

## Process

1. **Discovery**: Stakeholder interviews, current state assessment, data inventory
2. **Design**: Solution architecture, technology selection, implementation roadmap
3. **Implementation**: Agile development, integration testing, deployment
4. **Transition**: Knowledge transfer, training, monitoring setup

---

## Past Work

**Financial Institution**
AI underwriting system. 40% reduction in loan processing time, 25% improvement in risk assessment accuracy.

**Hospital Network**
Patient flow optimization. 30% reduction in wait times, 15% increase in capacity utilization.

**Retailer**
Demand forecasting system. 22% reduction in inventory costs with improved availability.

**Manufacturer**
Quality control and predictive maintenance. 35% reduction in defects, 18% efficiency improvement.


AI CONSULTING

AI-powered meeting intelligence that transcribes, summarizes, and extracts action items automatically.

# EchoBrief

<div class="flex items-center justify-between">
  <h3 class="text-xl font-semibold">Meeting Intelligence Platform</h3>
  <a href="https://echobrief-client.vercel.app/welcome" target="_blank" rel="noopener noreferrer" class="sb-component sb-component-block sb-component-button sb-component-button-secondary sb-component-button-icon">
    <span class="mr-2">Launch</span>
    <svg class="fill-current h-5 w-5" viewBox="0 0 24 24">
      <path d="M7 7h8.586L5.293 17.293l1.414 1.414L17 8.414V17h2V5H7v2z"/>
    </svg>
  </a>
</div>

---

## What It Is

AI-powered meeting intelligence platform. Upload or record any meeting and get automatic transcription, executive summaries, and extracted action items. Chat with AI about your meetings to find quotes, generate follow-up emails, and surface insights.

---

## The Problem

Meeting follow-up consumes significant time:

- Manual note-taking distracts from active participation
- Important details get lost or forgotten
- Action items lack clear ownership and deadlines
- No searchable record of past discussions
- Hours spent writing summaries and follow-up emails

Teams spend more time documenting meetings than acting on them.

---

## The Solution

Automated meeting intelligence with AI analysis:

- **AI Transcription** - 99% accuracy across 50+ languages with speaker detection
- **Instant Summaries** - Executive summaries generated in seconds
- **Action Extraction** - Tasks, deadlines, and assignees identified automatically
- **AI Chat** - Ask questions about any meeting, generate emails, find specific quotes

---

## How It Works

1. Upload audio/video file or record directly in browser
2. AI processes and transcribes the meeting (typically under 5 minutes)
3. Review generated summary and extracted action items
4. Chat with AI to query meeting content
5. Share insights and assign tasks to team

No manual transcription. No note-taking during meetings. Full searchable archive of all discussions.

---

## Technical Architecture

### Transcription Engine

- Advanced speech-to-text with 99%+ accuracy
- Multi-speaker diarization and identification
- Support for 50+ languages
- Handles various audio quality levels

### AI Analysis Layer

- Large language model processing for summarization
- Named entity recognition for people, dates, and tasks
- Sentiment analysis and topic clustering
- Context-aware action item extraction

### Document Analysis

- PDF and document upload support
- Contract and report analysis
- Chat interface for document queries
- Cross-reference with meeting content

### Integration Layer

- Zoom, Microsoft Teams, Google Meet support
- Slack notifications and summaries
- Salesforce and HubSpot CRM sync
- API access for custom integrations

---

## Features

| Feature | Description |
|---------|-------------|
| AI Transcription | 99% accuracy, 50+ languages, speaker detection |
| Instant Summaries | Executive summaries generated in seconds |
| Action Items | Automatic extraction with assignees and deadlines |
| AI Chat | Query any meeting - find quotes, generate emails |
| Document Analysis | Upload PDFs and contracts, chat to extract insights |
| Multi-Platform | Integrates with Zoom, Teams, Meet, Slack |

---

## Security Model

EchoBrief handles sensitive meeting content with appropriate security measures.

**Data Protection:**

- All data encrypted at rest and in transit
- SOC 2 compliant infrastructure
- User data not used for AI model training
- Full data deletion available on request

**Access Control:**

- Team-based permissions
- SSO support for enterprise
- Audit logging for compliance
- Role-based access controls

---

## Pricing

| Plan | Meetings | Processing | Price |
|------|----------|------------|-------|
| Free | 5/month | 60 minutes | $0 |
| Pro | 50/month | 600 minutes | $29/month |
| Enterprise | 1000/month | 10,000 minutes | $99/month |

All plans include AI transcription, summaries, and action item extraction. Pro and Enterprise include unlimited AI Chat and priority processing.

---


## Development Status

| Milestone | Target | Status |
|-----------|--------|--------|
| Core transcription engine | Q3 2025 | Live |
| AI summarization | Q3 2025 | Live |
| Action item extraction | Q3 2025 | Live |
| AI Chat interface | Q4 2025 | Live |
| Document analysis | Q1 2026 | Live |
| Slack integration | Q1 2026 | Live |
| Mobile application | Q2 2026 | Planned |
| Real-time transcription | Q2 2026 | Planned |
| Custom AI training | Q3 2026 | Planned |


ECHOBRIEF

Revenue streams across Portal Foundation products and services.


# Revenue Model

## Overview

Portal Foundation generates revenue through subscriptions, transaction fees, consulting, and data licensing.

---

## Revenue Streams

**Subscriptions**
Monthly and annual plans across products:
- CFA Learning Platform
- Market InfoAI
- AI Agent Platform
- AI Web Tools

Pricing: $25-50/month for individual plans, $499+/month for enterprise.

**Transaction Fees**
0.1% fee on swaps through Portal DEX aggregator.

**Consulting**
Project-based and retained consulting for AI integration. Typical engagements range from proof-of-concept to full implementation.

**Data Licensing**
Aggregated analytics on market activity and user behavior, licensed to institutions and research firms.

**White-Label**
Enterprise licensing of Portal products under client branding.

---

## Token Utility

The native token provides:
- Fee discounts on platform services
- Access to premium features
- Staking rewards from platform revenue share

---

## Product Revenue Mapping

| Product | Revenue Type |
|---------|--------------|
| CFA Learning Platform | Subscription |
| AI Agent Platform | Subscription |
| AI Web Tools | Subscription, usage-based |
| Portal DEX | Transaction fees |
| Telegram Bots | Subscription, white-label |
| Consulting | Project fees, retainers |

---

## Notes

Revenue projections depend on user adoption and market conditions. This document describes the revenue model structure, not guaranteed outcomes.

REVENUE MODEL

Trading terminal for perpetual futures with AI analysis and self-custody.

# Portal Terminal

<div class="flex items-center justify-between">
  <h3 class="text-xl font-semibold">Open Terminal</h3>
  <a href="https://trade.portalfoundation.ai" target="_blank" rel="noopener noreferrer" class="sb-component sb-component-block sb-component-button sb-component-button-secondary sb-component-button-icon">
    <span class="mr-2">Launch</span>
    <svg class="fill-current h-5 w-5" viewBox="0 0 24 24">
      <path d="M7 7h8.586L5.293 17.293l1.414 1.414L17 8.414V17h2V5H7v2z"/>
    </svg>
  </a>
</div>

---

## What It Is

Trading terminal for perpetual futures on decentralized infrastructure. All trades execute on-chain with full self-custody. Integrates AI-powered market analysis, real-time orderbook visualization, and one-click position management.

---

## The Problem

Perpetual futures trading is dominated by centralized exchanges with:
- Opaque order matching and liquidation practices
- Recurring security breaches and exchange failures
- Billions lost to hacks, fraud, and mismanagement

Decentralized alternatives exist but suffer from poor UX, limited tooling, and fragmented liquidity. Professional traders must choose between security (self-custody) and capability (advanced trading tools).

---

## The Solution

Frontend terminal built on decentralized trading infrastructure:
- On-chain execution with sub-second finality
- No deposits, no counterparty risk, no withdrawal delays
- AI copilot for market analysis integrated into workflow
- Professional-grade tools without sacrificing self-custody

---

## How It Works

1. Connect wallet
2. Select market and leverage
3. Place order (market or limit with TP/SL)
4. Trade executes on-chain
5. Manage positions from dashboard

No account creation. No KYC. Funds remain in wallet until trade execution.

---

## Technical Architecture

**Execution Layer**
- On-chain order matching with sub-second finality
- EIP-712 signing for gasless order placement
- Cross-margin and isolated margin modes
- Up to 50x leverage on major pairs

**AI Layer**
- GPT-4 powered market analysis
- Context-aware strategy generation
- Chart pattern recognition
- Natural language trade execution (planned)

**Analytics Layer**
- Cross-market funding rate dashboard
- Open interest heatmaps with signal detection
- Historical performance tracking
- Returns visualization (daily/weekly/monthly)

**Yield Layer**
- Morpho Protocol vault integration
- Direct deposit/withdraw from terminal
- APY comparison across lending markets

---

## Features

| Feature | Description |
|---------|-------------|
| AI Copilot | Ask questions about markets, get analysis with chart context |
| One-Click Execution | Market/limit orders with integrated TP/SL, leverage slider |
| Real-Time Orderbook | Depth visualization, large order detection, click-to-set-price |
| Open Interest Signals | Cross-market OI analysis, accumulation patterns, divergence detection |
| Funding Rate Dashboard | Live rates, annualized APR, historical charts, arbitrage opportunities |
| Strategy Templates | Pre-built strategies with customizable parameters |

---

## Security Model

Portal Terminal is a frontend. It does not custody assets, relay transactions, or operate trading infrastructure.

**What the system does NOT do:**
- Does not store API keys
- Does not maintain backend databases
- Does not create user accounts
- Does not hold or custody funds
- Does not add trust assumptions beyond the underlying protocol

**What happens:**
- All transactions signed locally in user's wallet
- Trading executes on decentralized, audited smart contracts
- Funds remain in wallet until trade execution
- Withdraw anytime without lockups, delays, or permissions

---

## Market Data

| Metric | Value |
|--------|-------|
| Daily perp futures volume | $75B+ |
| Total open interest | $45B+ |
| On decentralized venues | <5% |

The constraint for decentralized perpetuals has been UX, not technology.

---

## Development Status

| Milestone | Target | Status |
|-----------|--------|--------|
| Core trading terminal | Q4 2025 | Live |
| AI copilot integration | Q4 2025 | Live |
| Strategy templates | Q4 2025 | Live |
| Funding rate dashboard | Q1 2026 | Live |
| Open interest heatmap | Q1 2026 | Live |
| Morpho yield integration | Q1 2026 | Live |
| Copy trading system | Q2 2026 | In Progress |
| Natural language trade execution | Q2 2026 | Planned |
| Liquidation cascade alerts | Q2 2026 | Planned |
| Mobile application | Q3 2026 | Planned |
| Multi-chain expansion | Q3 2026 | Planned |


PORTAL TERMINAL

MCP server for smart contract generation, compilation, and deployment via Claude.

# PortalMCP

## What It Is

PortalMCP is a Model Context Protocol server that exposes smart contract tooling to Claude and other LLMs. Users describe what they want in natural language. Claude calls MCP endpoints to generate Solidity, compile it, and prepare deployment transactions. The user signs with their own wallet.

No private keys are stored or transmitted. All contract interactions require explicit user signature.

---

## The Problem

Smart contract development requires:
- Solidity proficiency
- Familiarity with compilation toolchains (Hardhat, Foundry)
- Understanding of deployment scripts and gas estimation
- Manual transaction construction

Most users cannot deploy a token or NFT collection without significant technical setup.

---

## The Solution

PortalMCP abstracts the development stack into MCP tool calls:

1. User prompts Claude: "Deploy an ERC-20 token called TestToken with 1M supply"
2. Claude calls `eth_generate_contract` to produce Solidity
3. Claude calls `eth_compile_contract` to get ABI and bytecode
4. Claude calls `eth_deploy_contract` to prepare the deployment transaction
5. User signs the transaction in their wallet
6. Contract is deployed

The same flow works for NFT collections, staking contracts, and custom Solidity.

---

## Available Tools

**Contract Pipeline**
| Tool | Function |
|------|----------|
| `eth_generate_contract` | Generate Solidity from specification |
| `eth_compile_contract` | Compile to ABI + bytecode |
| `eth_deploy_contract` | Prepare deployment transaction |
| `eth_call_contract` | Read contract state |
| `eth_send_transaction` | Prepare write transaction |

**Token Operations**
| Tool | Function |
|------|----------|
| `eth_create_token` | Generate ERC-20 contract |
| `eth_get_token_balance` | Query token balance |
| `eth_transfer_token` | Prepare transfer transaction |

**NFT Operations**
| Tool | Function |
|------|----------|
| `eth_create_nft_collection` | Generate ERC-721 contract |
| `eth_mint_nft` | Prepare mint transaction |
| `eth_get_nft_owner` | Query NFT ownership |

**DeFi Operations**
| Tool | Function |
|------|----------|
| `eth_create_staking_contract` | Generate staking contract |
| `eth_stake_tokens` | Prepare stake transaction |
| `eth_swap_tokens` | Prepare swap via DEX |

**Frontend**
| Tool | Function |
|------|----------|
| `generate_frontend` | Create React UI for contract |
| `deploy_dapp` | Deploy frontend to hosting |

---

## Security Model

PortalMCP is a tool server. It does not:
- Store private keys
- Sign transactions
- Hold custody of assets
- Execute transactions without user approval

All transactions are prepared and returned to the user for signing. Pre-signature simulation runs before any state-changing call.

---

## Supported Networks

- Ethereum Mainnet
- Sepolia Testnet
- Base
- Arbitrum
- Optimism
- Polygon

Network selection via `--network` flag at server startup.

---

## Development Status

| Milestone | Target | Status |
|-----------|--------|--------|
| Core tooling (ERC-20, NFT) | Q2 2025 | Complete |
| Frontend generator | Q3 2025 | Complete |
| Agent autonomy and looping | Q3 2025 | Planned |
| dApp templates | Q4 2025 | Planned |
| Multi-agent coordination | Q1 2026 | Planned |

---

## Technology Stack

- FastAPI + Node.js for MCP endpoints
- Solidity compilation via Foundry
- Ethers.js for transaction preparation
- Claude API for natural language processing

---

## Installation

**Requirements**
- Node.js v16+
- Ethereum wallet
- Claude API access

**Setup**

```bash
git clone https://github.com/portal-foundation/portalmcp.git
cd portalmcp
npm install
cp .env.example .env
# Add ANTHROPIC_API_KEY and INFURA_API_KEY to .env
npm run build
npm start
```

**Claude Configuration**

Add to `claude_desktop_config.json`:

```json
{
  "mcpServers": {
    "portalmcp": {
      "command": "node",
      "args": ["path/to/portalmcp/dist/index.js", "--network", "sepolia"]
    }
  }
}
```


PORTAL MCP

Non-custodial interface for Circle's CCTP v2. Native USDC transfers across eleven EVM networks with configurable finality.

# PortalCCTP

<div class="flex items-center justify-between">
  <h3 class="text-xl font-semibold">Launch Application</h3>
  <a href="https://send.portalfoundation.ai" target="_blank" rel="noopener noreferrer" class="sb-component sb-component-block sb-component-button sb-component-button-secondary sb-component-button-icon">
    <span class="mr-2">Open PortalCCTP</span>
    <svg class="fill-current h-5 w-5" viewBox="0 0 24 24">
      <path d="M7 7h8.586L5.293 17.293l1.414 1.414L17 8.414V17h2V5H7v2z"/>
    </svg>
  </a>
</div>

---

## The Problem

Cross-chain stablecoin movement is broken. Users face a fragmented landscape of third-party bridges, each with different security assumptions, liquidity constraints, and fee structures. Wrapped assets introduce counterparty risk. Liquidity pools can be drained. Settlement times vary from minutes to hours with no visibility into status.

Moving $10,000 in USDC from Arbitrum to Base should not require trusting an intermediary or accepting a synthetic asset.

---

## The Solution

PortalCCTP builds directly on Circle's Cross-Chain Transfer Protocol v2. This is the only mechanism where USDC is burned on the source chain and minted natively on the destination. No wrapped tokens. No liquidity pools. No third-party custody.

The protocol flow is straightforward: burn USDC on Chain A, wait for Circle's attestation service to sign the message, mint USDC on Chain B. PortalCCTP abstracts these three steps into a single interface with real-time status tracking.

---

## Technical Architecture

**Protocol Layer**
- Circle CCTP v2 MessageTransmitter and TokenMessenger contracts
- Circle Iris attestation service for cross-chain message verification
- Native burn/mint mechanism (no wrapped or synthetic assets)

**Application Layer**
- Non-custodial wallet integration (user signs all transactions)
- Automatic network detection and switching prompts
- Transaction indexing for burn/mint correlation

**Finality Options**
- *Fast Mode*: Pays priority fee to Circle's relayer network. Settlement in seconds.
- *Standard Mode*: Waits for natural attestation. Settlement in 10-15 minutes. No additional cost.

---

## Supported Networks

Eleven mainnet deployments:

| Network | Type | CCTP Version |
|---------|------|--------------|
| Ethereum | L1 | v2 |
| Avalanche | L1 | v2 |
| Arbitrum | L2 | v2 |
| Optimism | L2 | v2 |
| Base | L2 | v2 |
| Linea | L2 | v2 |
| Unichain | L2 | v2 |
| Sonic | L1 | v2 |
| World Chain | L2 | v2 |
| Sei | L1 | v2 |
| HyperEVM | L1 | v2 |

Additional networks added as Circle expands CCTP support.

---

## Advanced Configuration

For programmatic integrations and power users:

- **Destination Caller**: Restrict mint execution to a specific address. Required for mint-and-call patterns.
- **Hook Data**: Attach arbitrary hex payload to the mint transaction. Enables downstream contract execution.
- **Max Fee Override**: Set explicit ceiling on fast-mode priority fees.

---

## Security Model

PortalCCTP is a frontend. It does not custody assets, relay transactions, or operate infrastructure.

- All transactions signed locally in user's wallet
- Contract interactions limited to Circle's audited CCTP deployments
- Attestation verification handled by Circle's Iris service
- No backend servers, no API keys, no stored user data

The security boundary is Circle's protocol and the user's wallet. PortalCCTP adds no additional trust assumptions.

---

## Development Status

| Milestone | Target | Status |
|-----------|--------|--------|
| Burn/mint on eleven networks | Q3 2025 | ✓ Live |
| Fast finality integration | Q3 2025 | ✓ Live |
| Destination caller + hook fields | Q4 2025 | ✓ Live |
| Re-attestation recovery flow | Q4 2025 | In Progress |
| Attestation timing analytics | Q1 2026 | Planned |
| EURC support | Q2 2026 | Planned |

---

## Usage

1. Connect wallet
2. Select source chain, destination chain, amount
3. Select finality mode (fast/standard)
4. Approve USDC spend (first transfer only)
5. Execute burn transaction
6. Wait for attestation (status displayed in UI)
7. Switch to destination network
8. Execute mint transaction

Both transaction hashes displayed for verification.


PORTALCCTP

AI image and video generation with x402 micropayments. Pay per generation.


# Portal Create

<div class="flex items-center justify-between">
  <h3 class="text-xl font-semibold">Open Application</h3>
  <a href="https://gen.portalfoundation.ai" target="_blank" rel="noopener noreferrer" class="sb-component sb-component-block sb-component-button sb-component-button-secondary sb-component-button-icon">
    <span class="mr-2">Launch</span>
    <svg class="fill-current h-5 w-5" viewBox="0 0 24 24">
      <path d="M7 7h8.586L5.293 17.293l1.414 1.414L17 8.414V17h2V5H7v2z"/>
    </svg>
  </a>
</div>

---

## What It Is

AI content generation platform with pay-per-use pricing via x402 micropayments. Generate images and videos using multiple AI models. Pay in USDC on Base. No subscriptions.

---

## The Problem

AI generation platforms typically require:
- Monthly subscriptions with unused credits
- Prepaid credit systems
- No transparency on per-generation costs

---

## The Solution

Pay-per-generation via x402:
- Connect wallet (USDC on Base required)
- Select model and enter prompt
- Sign payment authorization (no gas fees)
- Content generates immediately

x402 uses EIP-3009 signatures for gas-free micropayments.

---

## How It Works

1. Connect wallet
2. Select AI model
3. Enter text prompt
4. Review price
5. Sign payment authorization
6. Download generated content

---

## x402 Payment Protocol

x402 extends HTTP 402 "Payment Required" for micropayments:

1. API returns 402 with payment requirements
2. Wallet creates EIP-3009 signature authorizing USDC transfer
3. CDP facilitator (Coinbase) verifies signature
4. Payment settles, content generates
5. No gas fees (signature, not transaction)

---

## Available Models

**Image Generation**
- FLUX (Schnell, Dev, Pro, Pro Ultra)
- Stable Diffusion (SDXL, SD 1.5)
- Recraft V3

**Video Generation**
- Kling (1.0, 1.6, Pro, 2.0)
- Google Veo 3
- Seedance (Lite, Pro)

**Utilities**
- Face restoration
- 2x upscaling
- Background removal

---

## Pricing

### Image Generation

| Model | Quality | Speed | Cost per Image |
|-------|---------|-------|----------------|
| FLUX Schnell | Good | Fastest | $0.002 |
| FLUX Dev | Great | Fast | $0.030 |
| FLUX Pro | Excellent | Medium | $0.040 |
| FLUX Pro 1.1 | Superior | Medium | $0.055 |
| FLUX Pro Ultra | Best | Slower | $0.080 |
| SDXL | High | Fast | $0.004 |
| SD 1.5 | Classic | Fast | $0.012 |
| Recraft V3 | Design | Medium | $0.055 |

### Video Generation

| Model | Resolution | Duration | Cost per Video |
|-------|-----------|----------|----------------|
| Kling 1.0 | 720p | 5s | $0.19 |
| Kling 1.6 | 720p | 5s | $0.32 |
| Kling Pro | 1080p | 5s | $0.71 |
| Kling 2.0 | 1080p | 10s | $1.75 |
| Veo 3 | 720p | 5s | $3.13 |
| Veo 3 Fast | 720p | 5s | $1.56 |
| Seedance Lite | 1080p | 5s | $0.26 |
| Seedance Pro | 1080p | 5s | $0.65 |

### Utility Tools

| Tool | Function | Cost |
|------|----------|------|
| Face Restore | Enhance faces | $0.0015 |
| Upscale 2x | Enlarge images | $0.0020 |
| Background Removal | Remove backgrounds | $0.0040 |

---

## Technology Stack

- Next.js 15, React 18, TypeScript
- RainbowKit, Wagmi, Viem for Web3
- x402 protocol with EIP-3009
- Prodia v2 API for generation
- Base network, USDC payments

---

## Development Status

| Phase | Focus | Status |
|-------|-------|--------|
| Phase 1 | Image generation, x402 | Live |
| Phase 2 | Video generation, utilities | Complete |
| Phase 3 | AI chat with micropayments | In Progress |
| Phase 4 | Custom model fine-tuning | Planned |
| Phase 5 | Multi-chain support | Planned |

---

## Links

- Platform: [gen.portalfoundation.ai](https://gen.portalfoundation.ai)
- x402 Protocol: [x402.org](https://x402.org)
- Base Network: [base.org](https://base.org)

PORTAL CREATE

Cross-chain token swaps across L1 and L2 networks.


# SwapCross

<div class="flex items-center justify-between">
  <h3 class="text-xl font-semibold">Open Application</h3>
  <a href="https://cross.portalfoundation.ai/" target="_blank" rel="noopener noreferrer" class="sb-component sb-component-block sb-component-button sb-component-button-secondary sb-component-button-icon">
    <span class="mr-2">Launch</span>
    <svg class="fill-current h-5 w-5" viewBox="0 0 24 24">
      <path d="M7 7h8.586L5.293 17.293l1.414 1.414L17 8.414V17h2V5H7v2z"/>
    </svg>
  </a>
</div>

---

## What It Is

Cross-chain swap interface. Select source chain, destination chain, input token, and output token. The application finds routes and executes the swap.

Non-custodial. Users sign transactions in their own wallet.

---

## How It Works

1. Connect wallet
2. Select source chain and token
3. Select destination chain and token
4. Enter amount
5. Review quote (rate, fees, estimated time)
6. Confirm transaction in wallet
7. Receive tokens on destination chain

---

## Supported Networks

| Network | Type |
|---------|------|
| Ethereum | L1 |
| BNB Chain | L1 |
| Polygon | L2 |
| Arbitrum | L2 |
| Optimism | L2 |
| Base | L2 |

Additional networks added as routing support expands.

---

## Supported Tokens

- Native assets: ETH, WETH, WBTC, MATIC, BNB
- Stablecoins: USDC, USDT, DAI
- DeFi tokens: varies by network

---

## Features

- Route optimization across multiple bridges
- Real-time quotes with slippage display
- Transaction status tracking
- Mobile-responsive interface

---

## Security

SwapCross is a frontend. It does not:
- Store private keys
- Hold custody of assets
- Execute transactions without user signature

All swaps require wallet approval.

---

## Development Status

| Phase | Focus | Status |
|-------|-------|--------|
| Phase 1 | Core swaps, UI | Live |
| Phase 2 | Expanded networks, improved quotes | In Progress |
| Phase 3 | History, alerts, preferences | Planned |


SWAPCROSS

Market data chatbot for crypto, stocks, and forex queries.


# PortalAI Chatbot

## What It Is

Chatbot for querying market data across crypto, stocks, and forex. Users ask questions in natural language. The bot fetches data from APIs, processes it, and returns formatted responses with analysis.

---

## How It Works

1. User asks question in natural language
2. Bot parses intent, asset, timeframe
3. Queries relevant APIs (Hyblock for crypto, StockForexProvider for equities/forex)
4. Processes and formats response
5. Returns human-readable answer with context

---

## Market Coverage

| Market | Data Available |
|--------|----------------|
| Crypto | Prices, market cap, volume, bid-ask, on-chain metrics |
| Stocks | Quotes, volume, dividends, earnings, technicals |
| Forex | Currency pairs, spreads, historical rates |

---

## Query Examples

- "What's the price of BTC?"
- "How has AAPL performed this week?"
- "What's the RSI for Ethereum?"
- "Compare BTC and ETH this month"
- "Is Bitcoin bullish or bearish?"

---

## Features

- Natural language query processing
- Multi-market data aggregation
- Technical indicator calculations
- Trend and sentiment analysis
- Formatted responses with key metrics

---

## Technical Stack

**Data Sources**
- Hyblock API (crypto)
- StockForexProvider API (stocks, forex)

**Processing**
- NLP for query parsing
- Financial entity recognition
- Metric calculation and formatting

---

## Development Status

| Phase | Focus | Status |
|-------|-------|--------|
| Phase 1 | Core data integration | Live |
| Phase 2 | Sentiment and predictive tools | In Progress |
| Phase 3 | Portfolio tracking, alerts | Planned |

PORTALAI CHATBOT

# Portal DEX

<div class="flex items-center justify-between">
  <h3 class="text-xl font-semibold">Open Application</h3>
  <a href="https://swap.portalfoundation.ai/" target="_blank" rel="noopener noreferrer" class="sb-component sb-component-block sb-component-button sb-component-button-secondary sb-component-button-icon">
    <span class="mr-2">Launch</span>
    <svg class="fill-current h-5 w-5" viewBox="0 0 24 24">
      <path d="M7 7h8.586L5.293 17.293l1.414 1.414L17 8.414V17h2V5H7v2z"/>
    </svg>
  </a>
</div>

---

## What It Is

Token swap interface with aggregation via 0x Protocol. Routes trades across multiple liquidity sources to find optimal pricing. Users connect wallet, select tokens, review quote, and execute swap.

---

## How It Works

1. Connect wallet
2. Select input and output tokens
3. Enter amount
4. Review quote (price, gas, slippage)
5. Execute swap

---

## Features

- 0x Protocol aggregation
- Multi-source liquidity routing
- Gas optimization
- Slippage protection
- Custom token import
- Real-time price feeds

---

## Fee Structure

| Fee Type | Amount |
|----------|--------|
| Swap fee | 0.1% |
| Protocol fee | 0.01% |

Token holders receive fee sharing via staking.

---

## Technical Stack

- 0x Protocol for routing
- ERC20 token support
- Web3 wallet integration
- Cross-chain compatibility (planned)

---

## Development Status

| Phase | Focus | Status |
|-------|-------|--------|
| Phase 1 | Core swap engine | Complete |
| Phase 2 | Cross-chain, analytics | In Progress |
| Phase 3 | Institutional tools | Planned |

PORTAL DEX

Telegram bot for trading, portfolio tracking, and price alerts via natural language.

# Portal Wallet

<div class="flex items-center justify-between">
  <h3 class="text-xl font-semibold">Open Bot</h3>
  <a href="https://t.me/PortalWallet_Bot" target="_blank" rel="noopener noreferrer" class="sb-component sb-component-block sb-component-button sb-component-button-secondary sb-component-button-icon">
    <span class="mr-2">Launch</span>
    <svg class="fill-current h-5 w-5" viewBox="0 0 24 24">
      <path d="M7 7h8.586L5.293 17.293l1.414 1.414L17 8.414V17h2V5H7v2z"/>
    </svg>
  </a>
</div>

---

## What It Is

Telegram bot for crypto trading and portfolio management. Users issue natural language commands to buy, sell, swap, set alerts, and track holdings. Trades route through 0x API with slippage and tax checks.

Wallet generation and storage handled within Telegram. Users can import existing wallets via mnemonic or private key.

---

## The Problem

Crypto trading typically requires:
- Separate wallet applications
- Manual token address input
- Understanding of DEX interfaces
- No integrated alerting or portfolio tracking

---

## The Solution

Single Telegram interface for:
- Natural language trade execution
- Price alerts with persistence
- Daily portfolio summaries
- Tax-aware routing via 0x

---

## How It Works

1. Generate or import wallet in Telegram
2. Issue command: "buy $50 of ETH" or "/alert PEPE > 0.000001"
3. Bot parses intent and fetches quotes from 0x
4. Slippage, tax, and approval checks run
5. Trade executes, confirmation sent
6. Portfolio and alerts update in background

---

## Commands

| Command | Function |
|---------|----------|
| Natural language | "buy $100 of ETH", "sell half my PEPE" |
| `/alert TOKEN CONDITION PRICE` | Set price alert |
| `/portfolio` | View holdings and values |
| `/subscribe` | Daily portfolio updates at 9AM |
| `/unsubscribe` | Stop daily updates |

---

## Features

- Natural language trade parsing
- 0x Permit2 routing with tax detection
- Encrypted wallet storage
- Contract interaction previews
- Price impact warnings
- Persistent alerts (survives restarts)

---

## Technical Stack

- Telegram Bot API
- 0x v2 API (Permit2)
- CoinGecko and DexScreener for pricing
- PostgreSQL for alerts and portfolio data
- NLP for command parsing

---

## Development Status

| Milestone | Target | Status |
|-----------|--------|--------|
| Natural language trading | Jul 2025 | Live |
| Persistent alerts | Jul 2025 | Live |
| Daily portfolio summaries | Jul 2025 | Live |
| Privacy swaps | Sep 2025 | In Progress |
| Sell-on-percentage-gain | Sep 2025 | In Progress |
| Limit orders | Sep 2025 | Planned |
| Portfolio rebalancing | Q4 2025 | Planned |

PORTAL WALLET

AI agents for DeFi operations, contract analysis, and yield optimization.

# Portal Agents Marketplace

<div class="flex items-center justify-between">
  <h3 class="text-xl font-semibold">Open Marketplace</h3>
  <a href="https://agents.portalfoundation.ai/" target="_blank" rel="noopener noreferrer" class="sb-component sb-component-block sb-component-button sb-component-button-secondary sb-component-button-icon">
    <span class="mr-2">Launch</span>
    <svg class="fill-current h-5 w-5" viewBox="0 0 24 24">
      <path d="M7 7h8.586L5.293 17.293l1.414 1.414L17 8.414V17h2V5H7v2z"/>
    </svg>
  </a>
</div>

---

## What It Is

Portal Agents Marketplace hosts AI agents that execute blockchain operations via natural language. Users describe what they want. Agents analyze options, propose actions, and execute approved transactions.

Each agent specializes in a domain: DeFi transactions, contract analysis, yield farming, or trading. All agents are non-custodial. Users sign transactions in their own wallet.

---

## The Problem

Blockchain operations require:
- Protocol-specific knowledge for each DeFi platform
- Manual monitoring of yield rates across chains
- Transaction construction and gas optimization
- Contract analysis for security assessment
- Position management across multiple protocols

---

## The Solution

Agents handle the complexity:
- Natural language input for operation requests
- Multi-protocol analysis and comparison
- Transaction preparation with gas optimization
- User approval before execution
- Continuous monitoring of positions

---

## Live Agents

**DeFi Agent**
Executes swaps, lending, and borrowing across Aave, Compound, and Uniswap. Integrates with 0x for swap routing. Handles gas optimization and transaction batching.

**Smart Contract Reader**
Analyzes contract source code and transaction history. Provides security assessment, function documentation, and risk scoring. Uses Etherscan and Moralis APIs.

**Yield Harvester**
Scans yield opportunities across chains via DefiLlama. Compares APYs, assesses protocol risk, and prepares migration transactions. Supports Compound v3 and Aave v3.

---

## Planned Agents

| Agent | Function | Target |
|-------|----------|--------|
| Privacy Swap | Zero-knowledge swaps via Railgun | Q4 2025 |
| Cross Chain Swap | Multi-chain bridging with route optimization | Q4 2025 |
| Perps Trading | Leverage trading on Hyperliquid and dYdX | Q1 2026 |
| Whale Scout | Large wallet tracking and sentiment analysis | Q1 2026 |
| Governance Steward | DAO proposal monitoring and voting | Q1 2026 |
| LP Optimizer | Uniswap v3 and Balancer position management | Q1 2026 |

---

## Technical Architecture

**DeFi Agent**
- 0x Protocol for swap routing
- EIP-712 signature handling
- Multi-provider Web3 connectivity
- Token registry with address resolution

**Smart Contract Reader**
- Moralis API for blockchain indexing
- Etherscan for source verification
- OpenAI for code analysis
- Caching layer for metadata

**Yield Harvester**
- DefiLlama for yield aggregation (200+ protocols)
- Direct Compound v3 and Aave v3 integration
- Risk-adjusted yield calculations
- Protocol categorization

**Communication Framework**
- WebSocket for real-time updates
- Persistent conversation context
- Error recovery with fallbacks
- Client-side transaction signing

---

## Security Model

Agents do not:
- Store private keys
- Sign transactions
- Hold custody of assets
- Execute without user approval

All transactions require explicit user signature. Agent actions are logged and auditable. Emergency shutdown available for all agents.

---

## How to Use

1. Connect wallet
2. Select agent from marketplace
3. Describe operation in natural language
4. Review agent proposal
5. Approve and sign transaction
6. Monitor execution

---

## Development Status

| Milestone | Target | Status |
|-----------|--------|--------|
| Core agents (DeFi, Contract Reader, Yield) | Q3 2025 | Live |
| Privacy and cross-chain agents | Q4 2025 | In Progress |
| Trading and analytics agents | Q1 2026 | Planned |
| Community agent marketplace | Q2 2026 | Planned |
| Custom agent builder | Q2 2026 | Planned |

PORTAL AGENTS MARKETPLACE

DeFi portfolio automation via Instadapp Smart Accounts with batched transactions.

# DSA Portfolio Manager

<div class="flex items-center justify-between">
  <h3 class="text-xl font-semibold">Open Application</h3>
  <a href="https://app.portal-dsa.com" target="_blank" rel="noopener noreferrer" class="sb-component sb-component-block sb-component-button sb-component-button-secondary sb-component-button-icon">
    <span class="mr-2">Launch</span>
    <svg class="fill-current h-5 w-5" viewBox="0 0 24 24">
      <path d="M7 7h8.586L5.293 17.293l1.414 1.414L17 8.414V17h2V5H7v2z"/>
    </svg>
  </a>
</div>

---

## What It Is

DeFi portfolio manager built on Instadapp's DeFi Smart Accounts (DSA). Bundles multiple protocol interactions into single transactions ("spells"). Supports Aave, Compound, Fluid, Uniswap. Natural language input for strategy description.

---

## The Problem

DeFi management requires:
- Separate transactions for each protocol
- High gas costs for multi-step operations
- Manual monitoring of rates across protocols
- Complex liquidation risk management

---

## The Solution

Batched transactions via DSA:
- Bundle 20+ actions into single transaction
- Cross-protocol yield comparison
- Automated rebalancing
- Flash loan integration
- Liquidation protection

---

## How It Works

1. Create or connect DSA (proxy contract)
2. Describe strategy or select template
3. System generates spell (batched actions)
4. Review and approve
5. Single transaction executes across protocols
6. Dashboard shows positions and history

---

## Supported Protocols

| Protocol | Operations |
|----------|------------|
| Aave | Deposit, borrow, repay |
| Compound | Supply, borrow, claim |
| Fluid | Liquidity layer access |
| Uniswap | Swap, LP provision |
| Curve | Swap, LP provision |

---

## Spell Types

**Yield Optimization**
- Harvest and compound rewards
- Migrate to highest-yield protocol
- Rate arbitrage

**Risk Management**
- Liquidation protection via flash loan
- Dynamic LTV adjustment
- Emergency withdrawal

**Rebalancing**
- Target allocation maintenance
- Cross-protocol migration
- Profit taking

---

## Technical Architecture

- Instadapp DSA as portfolio hub
- Connector-based protocol integrations
- Gas-optimized spell batching
- Flash loan integration
- Cross-chain support (Ethereum, Polygon, Arbitrum)

---

## Security

- Self-custody via personal DSA
- On-chain spell verification
- Emergency spell cancellation
- Built on audited Instadapp infrastructure

---

## Development Status

| Milestone | Target | Status |
|-----------|--------|--------|
| Basic spell casting | Q3 2025 | Live |
| AI yield optimization | Q3 2025 | Live |
| Flash loan integration | Q4 2025 | In Progress |
| Cross-chain spells | Q4 2025 | In Progress |
| Liquidation protection | Q1 2026 | Planned |
| DAO treasury management | Q2 2026 | Planned |


DSA PORTFOLIO MANAGER

Infrastructure and tools for building SaaS products under the Portal ecosystem.

# SaaS Platform

## What It Is

Portal provides infrastructure for building and launching SaaS products. Builders get access to shared infrastructure, AI capabilities, and development tools. Products built on the platform share revenue with token holders.

---

## The Problem

Launching a SaaS product requires:
- Infrastructure setup and maintenance
- Payment processing integration
- User authentication systems
- Analytics and monitoring
- Significant upfront capital

---

## The Solution

Portal provides:
- Shared infrastructure for hosting and scaling
- Pre-built authentication and payment systems
- AI integration capabilities
- Development frameworks and tools
- Revenue sharing model with token holders

---

## How It Works

1. Builders propose and develop SaaS products
2. Products use Portal infrastructure and tools
3. Revenue from subscriptions and usage fees is collected
4. Revenue is shared between builders and token holders

---

## Current Products

Products built on Portal infrastructure:
- CFA Learning Platform
- Market InfoAI
- AI Web Tools
- Telegram AI Bots

---

## Revenue Model

Products generate revenue through:
- Monthly and annual subscriptions
- Usage-based fees
- Enterprise licensing
- Professional services

A portion of revenue is distributed to token holders via staking.


SAAS PLATFORM

Local AI command-line interface with Ollama and llama.cpp support.


# PortalOS

## What It Is

Local AI command-line interface. Runs LLMs (LLaMA, Mistral, Phi) via Ollama or llama.cpp on your machine. No cloud dependencies. Natural language commands for file operations, document processing, and system control.

---

## The Problem

Cloud-based AI tools:
- Send data to external servers
- Require internet connection
- Limited control over model behavior

Local alternatives are often fragmented or require significant setup.

---

## The Solution

Single CLI that:
- Runs entirely locally
- Supports multiple model backends
- Processes documents and files via natural language
- Executes system commands

---

## How It Works

1. Select model backend (Ollama or llama.cpp)
2. Launch PortalOS in terminal
3. Issue natural language commands
4. Processing happens locally
5. Output returned in terminal

---

## Capabilities

**Document Processing**
- PDF, Markdown, TXT parsing
- OCR via Tesseract
- Summarization
- Key point extraction

**File Operations**
- Natural language file search
- Semantic search across content
- Batch operations
- Folder organization

**System Control**
- Execute shell commands via text
- Monitor CPU, RAM, disk
- Task automation
- Background process management

---

## Model Backends

| Backend | Use Case |
|---------|----------|
| Ollama | Modern systems, 8GB+ RAM, easy model switching |
| llama.cpp | Resource-constrained devices, quantized models |
| Custom | Fine-tuned or domain-specific models |

Supported models: LLaMA, Mistral, Phi, and others.

---

## Technical Architecture

- Modular NLP pipeline
- Context memory management
- Plugin support
- Cross-platform (Linux, macOS, Windows)
- RAG for personal knowledge bases

---

## Development Status

| Milestone | Target | Status |
|-----------|--------|--------|
| Core CLI and file operations | Q2 2025 | Live |
| Ollama and llama.cpp integration | Q2 2025 | Live |
| Advanced document parsing, OCR | Q3 2025 | In Progress |
| Plugin system | Q3 2025 | In Progress |
| Personal RAG training | Q4 2025 | Planned |
| Multimodal support | Q1 2026 | Planned |



PORTALOS

Market intelligence aggregation with AI summarization and sentiment analysis.

# MarketInfo AI

## What It Is

Market intelligence platform that aggregates data from news, social media, YouTube, newsletters, and on-chain sources. AI processes content for summarization, sentiment scoring, and trend detection. Insights delivered via email, dashboard, or API.

---

## The Problem

Market research requires monitoring:
- News sites and RSS feeds
- Twitter and social platforms
- YouTube videos and podcasts
- On-chain activity
- Newsletters and Substacks

No single platform aggregates all sources with AI analysis.

---

## The Solution

Unified aggregation with AI processing:
- Multi-source data collection
- NLP summarization
- Sentiment scoring
- Trend detection
- Personalized delivery

---

## How It Works

1. Data collected from news, social, video, on-chain sources
2. AI processes content (NLP, sentiment, pattern recognition)
3. Content scored and prioritized by relevance
4. Insights delivered via email, alerts, dashboard, or API

---

## Data Sources

| Source | Data Type |
|--------|-----------|
| News APIs | Headlines, articles |
| Twitter/X | Social sentiment |
| YouTube | Video transcripts, summaries |
| Substack | Newsletter content |
| On-chain | Price, volume, events |
| DeFi feeds | Protocol data |

---

## Features

- Real-time news monitoring
- YouTube video summarization
- Sentiment scoring across platforms
- Custom alerts
- Weekly and daily newsletters
- API access

---

## Subscription Tiers

| Tier | Features |
|------|----------|
| Basic | Weekly newsletters, trend summary |
| Premium | Daily updates, custom alerts, historical data |
| Enterprise | Real-time alerts, API access, white-label |

---

## Technical Stack

**Data Collection**
- News APIs and RSS
- Social media streams
- YouTube transcription
- On-chain indexing

**AI Processing**
- NLP summarization
- Sentiment analysis
- Anomaly detection
- Pattern recognition

**Delivery**
- Email automation
- Push notifications
- Web dashboard
- REST API

---

## Development Status

| Milestone | Target | Status |
|-----------|--------|--------|
| Core aggregation and newsletters | Q2 2025 | Live |
| Advanced sentiment and social signals | Q3 2025 | In Progress |
| API access and alert customization | Q3 2025 | In Progress |
| Predictive analytics | Q4 2025 | Planned |
| Enterprise dashboards | Q4 2025 | Planned |


MARKETINFO AI

Stock analysis tool with pattern recognition and signal scoring.


# TrendAI Analytics

## What It Is

Stock analysis platform that combines technical indicators, pattern recognition, and sentiment data to generate trading signals. Outputs include entry/exit prices, stop-loss levels, and confidence scores.

---

## The Problem

Traders must:
- Interpret multiple technical indicators manually
- Filter noise in volatile markets
- Connect sentiment and fundamentals to price action
- Determine signal reliability

---

## The Solution

Automated analysis:
- Pattern recognition across timeframes
- Multi-factor signal scoring
- Entry, target, and stop-loss levels
- Confidence ratings
- Visual charts with annotations

---

## How It Works

1. Pull data from Tiingo API and sentiment feeds
2. Run pattern recognition and indicator analysis
3. Generate signals with confidence scores
4. Display charts with entry/exit levels
5. Available via CLI or web interface

---

## Technical Indicators

- Moving averages (20, 50, 200-day)
- RSI, MACD, Bollinger Bands
- Volume analysis
- Breakout detection

---

## Pattern Recognition

| Pattern | Detection |
|---------|-----------|
| Double top/bottom | Yes |
| Head and shoulders | Yes |
| RSI divergence | Yes |
| Golden/death cross | Yes |
| Higher lows/lower highs | Yes |

Each pattern receives a confidence score.

---

## Signal Output

- BUY, SELL, or NEUTRAL classification
- Entry price
- Target price
- Stop-loss level
- Signal rationale
- Confidence score

---

## Technical Stack

**Data**
- Tiingo API for price and fundamentals
- TextBlob for sentiment analysis

**Analysis**
- TrendAnalyzer for indicators
- PatternDetector for chart patterns
- SignalGenerator for scoring

**Visualization**
- Plotly for charts
- Candlestick with volume overlays
- Pattern annotations

---

## Development Status

| Milestone | Target | Status |
|-----------|--------|--------|
| Core analysis engine | Q2 2025 | Complete |
| Multi-factor scoring | Q2 2025 | Live |
| Charting and summaries | Q3 2025 | In Progress |
| Portfolio backtesting | Q4 2025 | Planned |
| Web platform | Q4 2025 | Planned |






TRENDAI ANALYTICS

Adaptive learning system for CFA exam preparation with AI-generated feedback.


# CFA Learning Platform

## What It Is

An adaptive learning system for CFA exam preparation. Users progress through topic levels, complete assessments in multiple formats, and receive AI-generated explanations for incorrect answers. The system tracks performance and adjusts difficulty based on demonstrated mastery.

---

## The Problem

CFA pass rates are below 50%. Common issues:
- No structured progression through prerequisites
- Static content that does not adapt to user performance
- Difficulty identifying knowledge gaps
- Limited feedback on practice questions

---

## The Solution

Structured, adaptive learning:
- Topic dependency mapping with prerequisite requirements
- Multiple assessment formats (MCQ, calculation, written, case study)
- AI-generated explanations for answers
- Performance tracking with gap identification
- Difficulty adjustment based on mastery

---

## How It Works

1. User selects topic or follows recommended path
2. System presents content with prerequisite checks
3. Assessment tests understanding (MCQ, calculation, or written)
4. AI analyzes responses and provides feedback
5. Performance determines next topic or retry
6. Mastery exams verify readiness for each level

---

## Assessment Types

| Type | Format | Pass Threshold |
|------|--------|----------------|
| Knowledge Check | 5-10 MCQs | 70% |
| Calculation | Numeric input with step validation | Partial credit |
| Written Response | 200-500 words, AI-scored rubric | Clarity and depth |
| Case Study | Multi-format, 45-60 min | Scenario integration |
| Level Mastery | 25-30 MCQs + cases, 90 min | 75% |

---

## Technical Architecture

**Learning Engine**
- Prerequisite tracking
- Topic dependency graph
- Adaptive difficulty
- Spaced repetition scheduling

**Assessment System**
- MCQ engine with randomization
- Calculation validator
- AI rubric for written responses
- Case study framework

**AI Layer**
- GPT/Claude for answer explanations
- NLP for written response analysis
- Knowledge gap prediction
- Study plan recommendations

**Infrastructure**
- React/Next.js frontend
- Node.js/Express APIs
- PostgreSQL and MongoDB
- Redis caching
- JWT authentication

---

## Development Status

| Milestone | Target | Status |
|-----------|--------|--------|
| Progressive learning system | Q2 2025 | Complete |
| Calculation and case study engine | Q3 2025 | In Progress |
| AI explanation layer | Q3 2025 | In Progress |
| Gamification and streaks | Q4 2025 | Planned |
| Custom learning paths | Q1 2026 | Planned |
| Institutional dashboard | Q2 2026 | Planned |

---

## Use Cases

**CFA Candidates**
Structured preparation for Levels I, II, and III with adaptive pacing.

**Self-Learners**
Custom topic selection with AI-guided study plans.

**Institutions**
Employee training with group analytics and custom content.




CFA LEARNING PLATFORM

Browser extension, chat platform, and image tools with wallet integration.


# AI Web Tools

## What It Is

A suite of browser-based AI tools: multi-agent chat platform, Chrome extension for page summarization, and image generation. All tools integrate with Portal Wallet for on-chain actions.

---

## The Problem

Web-based AI tools are fragmented:
- Chat interfaces separate from browser context
- No connection between AI tools and blockchain wallets
- Image generation disconnected from other workflows
- Multiple subscriptions for different capabilities

---

## The Solution

Unified tooling:
- Chat platform with multiple specialized agents
- Chrome extension for in-page AI interaction
- Image generation and analysis
- Wallet integration for on-chain actions
- Single subscription across all tools

---

## Tools

**Multi-Agent Chat Platform**
- Multiple agents for different domains (research, trading, writing)
- Conversation context preserved across sessions
- Model selection (GPT-4, Claude, Ollama)
- Wallet connection for transaction preparation

**Chrome Extension**
- Page summarization
- PDF and document analysis
- Highlight extraction
- Translation
- In-page AI queries

**Image Tools**
- Text-to-image generation (Stable Diffusion, DALL-E)
- Object detection
- Style transfer
- OCR and document processing

---

## Technical Stack

**Language Processing**
- GPT-4, Claude, Ollama integration
- Multi-agent orchestration
- Context persistence
- Web data integration

**Vision**
- Stable Diffusion and DALL-E
- Object recognition
- OCR pipeline
- Content moderation

**Infrastructure**
- Multi-region cloud deployment
- Encrypted data handling
- Enterprise SSO support
- API connectors for business systems

---

## Development Status

| Milestone | Target | Status |
|-----------|--------|--------|
| Multi-agent chat platform | Q2 2025 | Complete |
| Chrome extension | Q2 2025 | Complete |
| Image tools | Q3 2025 | In Progress |
| Business integration toolkit | Q3 2025 | Planned |
| API marketplace | Q4 2025 | Planned |

---

## Use Cases

**Research**
Summarize articles, extract key points, query documents.

**Content Creation**
Generate images, create outlines, translate content.

**Trading**
Query market data, prepare transactions, monitor positions.

**Business**
Connect to internal APIs, process documents, generate reports.



AI WEB TOOLS

Multi-agent trading system with strategy-specific agents and signal aggregation.


# AI Hedge Fund

## What It Is

A proof-of-concept multi-agent trading system. Each agent implements a distinct investment strategy (value, technical, sentiment). A central orchestrator aggregates signals, weights by confidence, and generates trade recommendations.

This is a research project, not a live fund.

---

## The Problem

Trading systems typically:
- Use single models or fixed rule sets
- Lack multi-perspective analysis
- Cannot explain decision rationale
- Slow to adapt to changing conditions

---

## The Solution

Multiple specialized agents:
- Each agent analyzes markets using its strategy
- Signals scored by confidence
- Central aggregation weights and compares
- Dashboard shows reasoning and consensus

---

## Agents

| Agent | Strategy |
|-------|----------|
| Value Agent | DCF valuation, comparables, margin of safety |
| Technical Agent | Trend detection, pattern recognition, momentum |
| Sentiment Agent | News NLP, social media analysis |
| Buffett Agent | Quality metrics, moat analysis, long-term value |
| Ackman Agent | Activist thesis, catalyst identification |

Each agent operates independently and produces confidence-scored signals.

---

## Architecture

**Agent Layer**
- Containerized microservices per agent
- REST APIs for inter-agent communication
- Custom logic per strategy

**Data Layer**
- Alpha Vantage and Yahoo Finance for market data
- NLP pipeline for news and social sentiment
- Historical fundamental datasets
- Redis caching

**Processing Layer**
- Signal scoring engine
- Cross-agent comparison
- Anomaly detection
- Decision audit trail

**Decision Engine**
- Trade simulation
- Position sizing based on confidence
- Portfolio constraint validation
- Performance feedback to agents

---

## Technology Stack

- Python, FastAPI, Docker
- Kafka for event streaming
- MongoDB and PostgreSQL
- LangChain and CrewAI for agent orchestration
- GPT-4, FinBERT, Mistral for reasoning

---

## Development Status

| Milestone | Target | Status |
|-----------|--------|--------|
| Core agent framework | Q2 2025 | Complete |
| Technical and valuation agents | Q2 2025 | Complete |
| Sentiment and strategy agents | Q3 2025 | In Progress |
| Signal dashboard | Q3 2025 | Planned |
| Portfolio simulation | Q4 2025 | Planned |
| Risk management engine | Q4 2025 | Planned |
| Live execution | Q1 2026 | Planned |

---

## Use Cases

**Signal Research**
Analyze agent consensus and individual signal performance.

**Strategy Testing**
Plug in custom agent logic and compare against existing agents.

**Education**
Understand multi-factor analysis through transparent reasoning.


AI HEDGE FUND

Configurable Telegram bots with market data, AI chat, and voice commands.

# Telegram AI Bots

## What It Is

A Telegram bot framework with built-in market data, AI conversation, and voice input. Bots can be deployed to any group, configured per-group, and cloned for white-label use.

The bot responds to commands for price data, chart analysis, news, and general AI queries. It supports text, voice, and image input.

---

## The Problem

Telegram bots are typically:
- Single-purpose (price bot OR chat bot, not both)
- Difficult to customize without code changes
- Lacking voice and image input
- Not designed for white-label deployment

---

## The Solution

A modular bot framework with:
- 100+ commands across market data, AI chat, and utilities
- Per-group configuration and context
- Voice input via Whisper
- Image analysis via vision models
- One-command cloning for white-label deployment

---

## Commands

**AI Interaction**
| Command | Function |
|---------|----------|
| `/ask` | Text query to AI |
| `/talk` | Conversational mode |
| `/speech` | Voice input |
| `/vision` | Image analysis |
| `/summary` | Summarize conversation or document |

**Market Data**
| Command | Function |
|---------|----------|
| `/price` | Current price for any asset |
| `/fundingrate` | Perpetual funding rates |
| `/liqLevels` | Liquidation level analysis |
| `/sentiment` | Market sentiment indicators |
| `/chartview` | Generate chart image |

**News and Research**
| Command | Function |
|---------|----------|
| `/newstoday` | Daily news summary |
| `/trendingnews` | Trending topics |
| `/claude_research` | Deep research query |

**Trading Analysis**
| Command | Function |
|---------|----------|
| `/liquidationHeatmap` | Liquidation zones |
| `/openInterestVolumeDelta` | OI and volume analysis |
| `/orderBookAnalysis` | Order book depth |
| `/market_regime` | Current market regime |

**Configuration**
| Command | Function |
|---------|----------|
| `/setcontext` | Set group-specific context |
| `/clonebot` | Create white-label copy |
| `/settings` | Bot preferences |

---

## How It Works

1. Add bot to Telegram group
2. Run `/loadbot` to initialize
3. Configure with `/setcontext` for group-specific behavior
4. Use commands or natural conversation
5. Clone with `/clonebot` for white-label deployment

---

## Technology Stack

- Python with Telegram Bot API
- OpenAI and Claude for language processing
- Whisper for voice transcription
- Vision models for image analysis
- MongoDB for context storage
- Market data APIs for price and analytics

---

## Development Status

| Milestone | Target | Status |
|-----------|--------|--------|
| Core bot architecture | Q2 2025 | Complete |
| Market analysis modules | Q2 2025 | Complete |
| Voice and image commands | Q3 2025 | Complete |
| Bot cloning and white-label | Q3 2025 | Planned |
| API builder for custom logic | Q4 2025 | Planned |
| Template marketplace | Q1 2026 | Planned |

---

## Use Cases

**Trading Groups**
Bot provides price alerts, chart analysis, and sentiment data on command.

**Community Support**
Bot answers questions using project documentation as context.

**Research**
Voice queries for market analysis and news summaries.

**White-Label**
Clone and rebrand for specific communities or products.







TELEGRAM AI BOTS

Zero-knowledge authenticated messaging with on-chain storage and end-to-end encryption.

# ZK Portal Chat

<div class="flex items-center justify-between">
  <h3 class="text-xl font-semibold">Portal Chat</h3>
  <a href="https://zk.portalfoundation.ai" target="_blank" rel="noopener noreferrer" class="sb-component sb-component-block sb-component-button sb-component-button-secondary sb-component-button-icon">
    <span class="mr-2">Launch</span>
    <svg class="fill-current h-5 w-5" viewBox="0 0 24 24">
      <path d="M7 7h8.586L5.293 17.293l1.414 1.414L17 8.414V17h2V5H7v2z"/>
    </svg>
  </a>
</div>

---

## What It Is

Private messaging protocol powered by ZK-SNARKs. Every message is cryptographically authenticated through zero-knowledge proofs, encrypted end-to-end, and stored permanently on-chain. No accounts, no servers, no metadata leaks.

---

## The Problem

Private communication on the internet relies on centralized infrastructure with:

- Server-side access to message content and metadata
- Single points of failure and censorship
- Identity systems controlled by corporations
- No cryptographic guarantees of sender authenticity

Existing encrypted messengers still require trust in a central operator. Blockchain messaging attempts exist but expose sender addresses and lack cryptographic privacy guarantees.

---

## The Solution

Wallet-native messaging with zero-knowledge authentication:

- **ZK-SNARK proofs** - Cryptographically prove identity without exposing keys
- **On-chain storage** - Messages persist on Ethereum and Base, censorship-resistant
- **Stealth addresses** - Recipient identity obscured on-chain
- **Wallet-derived keys** - No passwords, no seed phrases to manage

---

## How It Works

1. Connect wallet (Ethereum or Base)
2. Derive viewing key (one-time wallet signature)
3. Enter recipient wallet address
4. Compose and send message
5. ZK-SNARK proof generated in browser
6. Encrypted message posted on-chain with proof verification

No account creation. No server registration. Viewing keys deterministically derived from wallet signature - same key on any device.

---

## Technical Architecture

### Zero-Knowledge Layer

- **Groth16 ZK-SNARKs** for viewing key ownership proofs
- **Poseidon hashing** for ZK-optimized commitments and nullifiers
- **Client-side proof generation** via snarkjs in browser
- **On-chain verification** through deployed Verifier contracts

### ZK Circuit

- Proves ownership of viewing key without revealing it
- Generates nullifiers to prevent replay attacks
- Creates commitments binding message to sender
- All computation happens client-side - keys never transmitted

### Encryption Layer

- End-to-end encrypted message content
- EIP-5564 stealth addresses for recipient privacy
- Only intended recipient can decrypt

### Storage Layer

- Encrypted messages stored on Ethereum and Base
- ZKNoteRegistry smart contract with proof verification
- Permanent, immutable message history
- No server infrastructure required

### Key Management

- Viewing keys derived from wallet signatures
- Deterministic derivation - same key on any device with your wallet
- Keys never transmitted or stored remotely
- Sign once, decrypt forever

---

## Features

| Feature | Description |
|---------|-------------|
| ZK-SNARK Authentication | Every message authenticated via zero-knowledge proof |
| Stealth Recipients | On-chain addresses unlinkable to recipient identity |
| Browser Proof Generation | Groth16 proofs generated entirely client-side |
| Wallet-Native Identity | No usernames, no passwords - your wallet is your identity |
| On-Chain Persistence | Messages survive server failures, domain seizures, shutdowns |
| Multi-Chain Support | Live on Ethereum and Base |

---

## Security Model

ZK Portal Chat is a client application. It does not operate servers, store messages, or custody keys.

**What the system does NOT do:**

- Does not store viewing keys or private keys
- Does not maintain message databases
- Does not relay messages through servers
- Does not collect metadata or analytics
- Does not require trust in any operator

**What happens:**

- All encryption/decryption happens locally in browser
- ZK-SNARK proofs generated client-side using snarkjs
- Messages posted directly to smart contract with proof
- On-chain Verifier confirms proof validity
- Only recipients with correct viewing key can decrypt

---

## Privacy Guarantees

| Property | Guarantee |
|----------|-----------|
| Message Content | Encrypted - unreadable on-chain |
| Sender Identity | Proven via ZK-SNARK - address not linked to message |
| Recipient Identity | Stealth address - only recipient knows it's for them |
| Key Custody | Viewing keys never leave device |

---

## Deployed Contracts

| Network | Contract |
|---------|----------|
| Base | ViewingKeyOwnershipVerifier |
| Base | ZKNoteRegistry |
| Ethereum | ZKNoteRegistry |
| Ethereum | ViewingKeyOwnershipVerifier |

---

## Development Status

| Milestone | Target | Status |
|-----------|--------|--------|
| Core encrypted messaging | Q4 2025 | Live |
| Viewing key derivation | Q4 2025 | Live |
| On-chain message storage | Q4 2025 | Live |
| ZK-SNARK proof generation (Groth16) | Q1 2026 | Live |
| ZKNoteRegistry + Verifier contracts | Q1 2026 | Live |
| Multi-chain (Ethereum + Base) | Q1 2026 | Live |
| Stealth address integration | Q1 2026 | Live |
| Group messaging | Q2 2026 | Planned |
| Encrypted attachments | Q2 2026 | Planned |
| ENS/Basename resolution | Q2 2026 | Planned |
| Mobile application | Q3 2026 | Planned |


Projects