> For the complete documentation index, see [llms.txt](https://thirdfi.gitbook.io/thirdfi.org-1/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://thirdfi.gitbook.io/thirdfi.org-1/product/thirdfi-v2/technical-overview/network-architecture.md).
# Network Architecture
ThirdFi network is building a blockchain framework specifically tailored to support AI functionalities, focusing on high data throughput, low latency, and enhanced security. Here's how the architecture is structured:
### Core Blockchain Layer
* **Function**: This is the primary layer where all transactions are recorded. It supports basic blockchain functionalities, including token transactions, consensus mechanisms, and network governance.
* **Technology**: Uses a Proof of Stake (PoS) consensus algorithm designed for energy efficiency and fast transaction processing. This layer is built using a scalable blockchain framework that can handle high transaction volumes typical in AI applications.
### Computational Layer
* **Function**: Dedicated to processing intensive AI tasks. This layer handles off-chain computations to maintain network performance without overloading the main blockchain.
* **Technology**: Utilizes state channels for executing AI tasks related to the blockchain but off the main ledger, allowing for real-time processing without transaction fee implications. This layer is also where machine learning models can interact, learn, and store intermediate states without committing all data to the main blockchain.
### Data Management Layer
* **Function**: Focuses on managing the vast amounts of data required for AI operations, including storage, access, and privacy.
* **Technology**: Integrates decentralized storage solutions such as the InterPlanetary File System (IPFS) for distributing and storing large datasets securely. It uses advanced encryption methods, such as homomorphic encryption and zero-knowledge proofs, to allow computations on data without exposing the underlying data itself.
