Leveraging AI in The Graph Network
Sam Green and Tomasz Kornuta of Semiotic Labs, core developers of The Graph, have provided an overview of past, present, and future AI efforts in the network. The Graph is a decentralized protocol for indexing blockchain data, making it available for querying in various applications such as dapp frontends, plots, dashboards, and data analytics. By leveraging AI, The Graph aims to improve automation within the network and lower the barrier to entry for accessing its rich web3 data.
The Graph uses incentive mechanisms to encourage optimal and honest behavior among protocol participants. These incentives guide the behavior of Indexers, Curators, Delegators, and Fishermen. Decentralization results in participants needing to make complex decisions; Semiotic Labs uses AI to deploy tools simplifying this decision-making process. They have contributed to the development of two AI-related tools: AutoAgora and the Allocation Optimizer, which help Indexers improve their protocol performance and revenue.
AutoAgora: Automating Pricing on The Graph Network
The primary purpose of The Graph is to serve queries to users through interactions between Indexers (data sellers), consumers (data buyers), and gateways. When customers send a query to a gateway, it distributes the query among Indexers based on factors such as price-bids, quality of service (QoS), latency, etc. Indexers earn money by serving queries while controlling the prices they offer.
Indexers express their price-bids using a domain-specific language called Agora. However, creating and updating Agora models can be time-consuming, leading many Indexers to use static, flat pricing models instead. To help Indexers with pricing, Semiotic Labs created an open-source tool called AutoAgora. This tool automates the process of creating and updating Agora price models, making it easier for Indexers to offer dynamic pricing reflecting the actual cost of serving a particular query shape.
AutoAgora Modules and AI Integration
AutoAgora consists of several modules that work together to automate the creation and updating of Agora price models. These modules include:
- Logs Processing: parsing logs to extract incoming queries, their shapes, and execution times;
- Relative Cost Discovery: grouping similar query shapes and calculating their resource usage statistics (e.g., mean execution times);
- Absolute Price Discovery: maximizing revenue by adjusting prices based on past query volume served.
AI is used in the Absolute Price Discovery module, helping Indexers offer more competitive and flexible pricing for their query services on The Graph Network.
As AI continues to be integrated into The Graph Network, it will enable better decision-making processes and dynamic pricing strategies for participants, ultimately enhancing the overall efficiency and effectiveness of the network.