> For the complete documentation index, see [llms.txt](https://prismnetwork.gitbook.io/prismnetwork-docs/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://prismnetwork.gitbook.io/prismnetwork-docs/4.-technical-architecture/4.3-technology-stack-overview.md).
# 4.3 Technology Stack Overview
Prism Network's technology stack adopts a modular design, with each layer communicating through standardized interfaces, ensuring system scalability, maintainability, and high performance.
Account Layer serves as the entry point for users interacting with Prism Network. It is built on mainstream Web3 wallet SDKs, supporting connections and signature-based logins from major wallets such as MetaMask, OKX Wallet, and TokenPocket. At this layer, we introduce EIP-712 typed signatures or equivalent human-readable signature formats, enabling users to clearly see the authorization scope, execution scope, amount limits, validity period, and revocation conditions before authorizing a strategy account to execute trades. This design effectively reduces blind-signing risks while providing the system with comprehensive account permission auditing capabilities. The Account Layer also manages users' asset balances, strategy subscription status, order history, profit records, and execution logs, providing foundational data support for subsequent strategy operations and trade executions.
Strategy Layer is the core intelligence of Prism Network. We have chosen Python as the primary development language, combined with the Ray distributed computing framework, to build a flexible and scalable strategy execution environment. Python's rich AI and data analytics ecosystem (such as NumPy, Pandas, and PyTorch) enables us to rapidly implement signal processing, probability modeling, and strategy backtesting. The Ray framework handles the distributed orchestration of strategy tasks, supporting parallel execution of multiple strategies, state management, and fault recovery. Additionally, the Strategy Layer incorporates an experience learning module based on Reinforcement Learning — the system continuously optimizes strategy parameters, signal weights, and risk control thresholds based on historical execution results and event settlement feedback, enabling strategy accounts to evolve autonomously.
Execution Layer undertakes the critical responsibility of converting strategy instructions into actual trades. We have implemented the execution engine in Rust, whose high performance, memory safety, and concurrency features make it ideally suited for high-frequency order routing and real-time monitoring tasks. The Execution Layer communicates with the Strategy Layer via the gRPC protocol. Upon receiving standardized execution instructions, it completes order routing, order splitting, order placement, status monitoring, exception handling, and result feedback. gRPC's efficient binary serialization and bidirectional streaming capabilities ensure low-latency, high-throughput interactions between the Strategy Layer and Execution Layer. Within the Execution Layer, we have also implemented adapters for different prediction market types (order book markets and AMM markets), automatically selecting the optimal execution path based on price, depth, slippage, and fees.
Data Layer is responsible for providing the entire system with real-time, accurate, and structured external data inputs. We continuously obtain real-time event prices, order book depth, and trading volume data from prediction markets (Polymarket, Kalshi, etc.) via WebSocket connections, while simultaneously indexing on-chain event data using The Graph's subgraphs to ensure efficient querying and synchronization. For external signals such as news, social sentiment, on-chain fund flows, and macro data, the Data Layer performs parallel collection through multi-source APIs, completing preprocessing steps including data cleaning, deduplication, timestamp alignment, entity recognition, and event mapping. The processed data is written into a unified data structure for invocation by the Strategy Layer and signal engines.
Protocol Layer is Prism Network's on-chain settlement and governance foundation, built on Solidity smart contracts and running on EVM-compatible chains. The Protocol Layer defines standard interfaces for strategy accounts, profit calculation and distribution rules, node collaboration mechanisms, and DAO governance processes. To ensure computational integrity while protecting user privacy, the Protocol Layer introduces zero-knowledge proof (ZK-SNARK) technology — nodes can generate verifiable computation proofs after executing tasks, submitting them to verification contracts for result validation without exposing specific task details or data content. This mechanism provides Prism Network with a decentralized, verifiable, and privacy-preserving trust foundation, enabling developers, users, and nodes to complete computational settlement and profit distribution without the need for trusted intermediaries.
Through the coordinated operation of these five modular layers, Prism Network forms a complete closed loop from data input, signal analysis, strategy generation, and order execution to on-chain settlement, delivering a one-stop AI strategy account service for prediction markets.
---
# Agent Instructions
This documentation is published with GitBook. GitBook is the documentation platform designed so that both humans and AI agents can read, navigate, and reason over technical content effectively. Learn more at gitbook.com.
## Querying This Documentation
If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question.
Perform an HTTP GET request on the current page URL with the `ask` query parameter, and the optional `goal` query parameter:
```
GET https://prismnetwork.gitbook.io/prismnetwork-docs/4.-technical-architecture/4.3-technology-stack-overview.md?ask=&goal=
```
`ask` is the immediate question: it should be specific, self-contained, and written in natural language.
`goal` is optional and describes the broader end goal you are ultimately trying to accomplish on behalf of the user. GitBook uses it to tailor the answer towards what is most useful for that goal.
The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.
Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.