Kinds.ai Whitepaper
  • Introduction
    • The Story
    • Getting Started
  • The AI Agents
    • Customized Agents (Conversational Agents)
    • Functional Agents
  • Decentralized Agent Layer for Everyone
    • AI Agent Platform
      • Initial Agent Offering (IAO)
      • Create and Customize AI Agents
        • Bonding Curve & Graduated Token
        • AI Agent Deployment
      • Tokenize AI Agent
        • IAO Mechanism
        • Rug-free Mechanism
    • $KINDS Token Introduction
      • $KINDS Token-economics
      • $KINDS Distribution
  • Kinds Protocol
    • Multi-Agent Collaboration
      • Co-ownership of KINDS Agents
      • Decentralized Deployment of KINDS Agents
      • Business Model
    • Collaborative Input and Traceability
    • Leaderboard
  • Important Information
    • Roadmap
    • Official Links
Powered by GitBook
On this page
  1. Decentralized Agent Layer for Everyone
  2. AI Agent Platform
  3. Create and Customize AI Agents

AI Agent Deployment

Deploying an AI Agent in the Kinds.ai ecosystem involves a seamless integration of on-chain and off-chain processes. This ensures that the agent is fully operational and ready for interaction upon creation while allowing for future updates and customization.

Key Steps in AI Agent Deployment

  1. AI Model Selection

    • At the time of creation, a pre-defined cognitive model is assigned as the agent’s default core.

    • Users have the flexibility to modify or upload custom models to enhance their agent’s capabilities post-deployment.

  2. Character Card & AI Model Preparation

    • A character card and the agent's cognitive core are created using user-provided information during the creation process.

    • This step ensures that the agent’s personality and functionality are aligned with user specifications.

  3. Mint Default Contribution

    • To kickstart deployment, a default contribution for the agent’s cognitive model is minted, enabling immediate operability.

  4. Mint Contribution NFT

    • A Contribution NFT is minted and linked to the agent’s Token-Bound Address (TBA), serving as proof of the contribution to the AI model.

  5. Create and Approve Contribution Proposal

    • A proposal is generated, detailing the contributions to the agent’s AI model.

    • Using delegated voting power, the system or community approves the proposal, validating modifications to the agent.

  6. Mint Service NFT

    • A Service NFT is issued to finalize the integration, allowing the AI model updates to be incorporated into the agent’s functionality.

  7. Model Update and Deployment

    • The agent’s AI model is retrieved from IPFS and deployed into the Agent Runner, a specialized hosting environment that manages the agent’s Cognitive, Voice, and Visual cores.

Creation Statuses

The agent’s deployment progresses through real-time statuses, providing users with clear updates:

  • ACTIVATING: The agent is being minted on-chain while its off-chain deployment is finalized. This process typically takes less than five minutes.

  • AVAILABLE: Deployment is complete, and the agent becomes fully operational. Users can now interact with the agent via platforms like Telegram and customize its settings through the dashboard.

Decentralized Updates

Post-deployment, users can manage updates to the agent’s AI model or behavior through a decentralized contribution process. Changes are submitted as proposals and require community approval via delegated voting power to ensure transparency and accountability.

PreviousBonding Curve & Graduated TokenNextTokenize AI Agent

Last updated 2 months ago