Skip to content

AgenticPlaceholder/agentkit-biddder

 
 

Repository files navigation

AgenticPlaceholder Bidding Agent

System Architecture

Our bidding agent system utilizes a sophisticated multi-layered architecture that ensures efficient, automated, and reliable ad placement across the digital billboard network. The following diagram illustrates the system's key components and their interactions:

flowchart TB
    subgraph User["User Interface"]
        UI["Web Interface"]
        Mobile["Mobile App"]
    end

    subgraph AgentSystem["Bidding Agent System"]
        BA["Bidding Agent Core"]
        BS["Bidding Strategies"]
        AM["Auction Monitor"]
        PM["Price Monitor"]
        
        subgraph Strategy["Strategy Engine"]
            ML["Large Language Module"]
            OPT["Prompt Engineering Engine"]
        end
        
        BA --> BS
        BA --> AM
        BA --> PM
        BS --> Strategy
        Strategy --> ML
        Strategy --> OPT
        Strategy --> RISK
    end

    subgraph Blockchain["Blockchain Layer"]
        direction TB
        BSC["Base Network"]
        
        subgraph SmartContracts["Smart Contracts"]
            NFT["NFT Contract"]
            MP["Marketplace Contract"]
           
        end
        
        BSC --> SmartContracts
    end

    subgraph Display["Display Network"]
        RPI["Raspberry Pi Devices"]
        DISP["Digital Billboards"]
        
        RPI --> DISP
    end

    subgraph External["External Services"]
        CB["Coinbase Agent Kit"]
        TN["True Network"]
        IPFS["IPFS Storage"]
    end

    User --> BA
    BA --> BSC
    BA --> External
    SmartContracts --> RPI
    TN --> SmartContracts

    classDef blockchain fill:#f9f,stroke:#333,stroke-width:2px
    classDef agent fill:#bbf,stroke:#333,stroke-width:2px
    classDef external fill:#bfb,stroke:#333,stroke-width:2px
    classDef display fill:#fbb,stroke:#333,stroke-width:2px
    
    class BSC,SmartContracts blockchain
    class AgentSystem,BA,BS,Strategy agent
    class External,CB,TN external
    class Display,RPI,DISP display
Loading

Ad Image

Architecture Components

The system architecture consists of four main layers that work together seamlessly to provide automated ad bidding and placement:

  1. Users can chat with the Publisher Agent. It extracts the necessariy details from the user's conversation to generate a Title,Text and an Image Description. The image description is then used to generate an image with DaLLE. Afterwards this Ad is published as an NFT.

  2. Bidding Agent System serves as the intelligent core of our platform. The bidding agent continuously monitors market conditions, analyzes pricing trends, and executes optimal bidding strategies. It uses 4O-mini for prediction, risk management for exposure control, and an optimization engine for precise bid timing and pricing.

  3. Blockchain Layer, built on the Base network, manages all transactions and ad ownership through smart contracts. The NFT contract handles ad content ownership, while the marketplace contract manages the dutch auction mechanism for ad space allocation.

  4. Display Network consists of Raspberry Pi devices connected to physical digital billboards. These edge nodes receive ad content directly from the blockchain and ensure timely display updates.

Integration Points

The system leverages several external services for enhanced functionality:

  • Coinbase Agent Kit provides the foundational framework for our autonomous bidding agent
  • True Network's attestation layer ensures publisher accountability through reputation scoring
  • IPFS delivers decentralized storage for ad content and metadata

Data Flow

The architecture supports a streamlined data flow where:

  1. Users input their advertising requirements through the interface.
  2. The bidding agent processes these requirements and monitors market conditions
  3. Smart contracts handle the auction mechanics and ownership verification
  4. Display devices receive and showcase the winning advertisements
  5. True Network validates performance metrics and updates reputation scores

Releases

No releases published

Packages

No packages published

Languages

  • TypeScript 55.5%
  • Python 43.5%
  • Other 1.0%