Blog

Tangle Network Basics

These posts are about the Tangle Network protocol, a separate project from Tangle's infrastructure for AI agents. Introduction to Tangle covers the agent infrastructure.

The Tangle Network is a protocol where independent operators run services, such as model inference and code execution, coordinated by blockchain contracts. A Blueprint defines a service, a Service is one deployed instance of it, and a Job is one call to it; the computation runs on the operator's machines.

For builders and buyers deciding whether an operator-run service fits their workload. The posts keep three claims apart: that a Job ran, that it was paid for, and that its result is correct.

Start with when independent operators are worth their cost. Then follow one request through a Blueprint, see how results are checked, build and deploy a service, add model inference and code execution, and test what trusted hardware proves.

  1. 1 Why Decentralized AI Infrastructure?

    When choosing among independent operators earns its cost, and when a single cloud provider is the better choice.

  2. 2 How Blueprints Work

    Follow one request from a published Blueprint through operator registration, a Service, a Job call, and settlement.

  3. 3 How Decentralized AI Infrastructure Verifies Work

    Which check answers which claim: matching results, stake thresholds, challenge windows, trusted hardware, zero-knowledge proofs, and evaluations.

  4. 4 How to Build a Tangle Blueprint: Test and Deploy

    Turn a tested Rust function into an operator-run service, from a local job router to a testnet deployment.

  5. 5 AI Agent Infrastructure: Inference and Code Execution

    Separate model inference from generated-code execution, and check model identity, sandbox policy, and the result for each.

  6. 6 Secure Containers for AI Agents: What TEEs Can Prove

    What a trusted execution environment (TEE) proves: whether the host can read private work, and which program ran.