Chainsight is an interchain layer built for high-capacity data processing. It enables complex processes such as cross-chain messaging without relying on a single point of trust. This characteristic allows for the cost-effective development of products such as Dynamic Volatility Oracle (which continuously updates price volatility over a set period) or Cross-chain Lending (which hinges on cross-chain credit management). ^[1][2]

Overview

Launched in 2023, Chainsight provides a solution to the fragmentation of on-chain data across various blockchains, facilitating seamless access to historical data. Acting as a universal data extension layer, it enables users to retrieve information from any blockchain via smart contracts. This approach enhances transparency and reduces reliance on trust points. Chainsight offers access to a comprehensive dataset from the entire blockchain ecosystem. ^[3]

Chainsight is a toolkit for analyzing and showcasing intricate indicators and cross-chain data. Each piece of data stands independently as an open API, and by integrating these pieces effectively, they can create new data affordably and easily, thus reusing the wisdom of others. ^[4]

Utility

Chainsight is effective for any decentralized apps (Dapps) that handle "data" in a creative way. A typical example is Dynamic DeFi based on data-driven decision-making. When considering a Dynamic Money Market, a feature could be implemented through Chainsight's historical data that would automatically list assets there if they meet specified criteria.

Not just for DeFi, users can target and leverage any data, including data queries from Web2 social media APIs, sports/sports data for betting applications, etc.^[5]

Architecture

System Outline

Chainsight is based on Internet Computer's Chain-key Cryptography. This allows data to be kept in sync with any L1/ L2 in a trustless manner. Data is first indexed into the storage of canister smart contracts running on top of Internet Computer, and by connecting each canister smart contract like Lego blocks, more complex data can be handled at low cost. ^[6]

Components deployed by developers using the CLI or UI run on Internet Computer as Composable Data Oracles. Apart from these, Chainsight Management Canisters have been deployed to control the entire network, and incentives for data providers and alerts for corrupt data will be implemented in the future milestone. ^[6]

On-Chain Data Flow

Chainsight eliminates the need for off-chain setups and establishes an efficient on-chain data processing pipeline. To achieve this, users need to select the data source and define some basic logic to transform the raw data into useful information. This on-chain approach reduces the reliance on trust points and provides a more cost-effective solution for utilizing on-chain data in applications. ^[7]

Collecting Raw Data across On-Chain

Chainsight collects arbitrary data from arbitrary blockchains. There are two main types of data: event data, as represented by Transfer in ERC-20, and data that is periodically retrieved from a view function for the current value. HTTP calls are periodically made to the blockchain nodes to update Chainsight's data to the latest state. The interval to be retrieved can be specified by the Indexer owner. Once the data is indexed on the chain site, anyone can use it as a data source. Data can be combined and customized freely for efficient and meaningful on-chain data.

Threshold ECDSA

Data on Chainsight can be taken out to other blockchains. ECDSA signing can be performed without reconstructing the private key using secret shares generated on multiple distributed nodes. This allows for immediate transactions with decentralized key management when writing one piece of data to another blockchain.^[7]

Data Processing Components

Chainsight allows developers to access and utilize desired on-chain data through various data processing components. The system comprises an Indexer that collects on-chain data and stores it, an Algorithm Lens that extracts and references data from the Indexer, and a Relayer that writes the data to other blockchains. ^[8]

Together, these components enable developers to efficiently collect and use a wide range of on-chain data. Indexers can be classified into three types based on their roles, resulting in a total of five components, including the Algorithm Lens and Relayer. ^[8]

Event Indexer

Event Indexer synchronizes event data from a specified blockchain and stores a list of event records. For example, collecting ERC-20 Transfer events can provide a basis for analyzing token transfer history. This data is primarily prepared for use with subsequent components.

Snapshot Indexer

Snapshot Indexer periodically queries view functions, such as Smart Contracts and Algorithm Lens, to track results. Periodic snapshot data can serve as historical data for further analysis.

Algorithm Indexer

The algorithm Indexer processes the raw data from Event Indexer and Snapshot Indexer for analysis and generates insightful metrics. For instance, the Algorithm Indexer takes the raw data of ERC-20 Transfer events from the Event Indexer and creates separate mapping data on how many tokens are held by each address. ^[8]

In contrast to the Indexer, the Algorithm Lens and Relayer do not store the state themselves but write to other blockchains.

Algorithm Lens

Retrieves existing data from the Indexer using custom logic, enabling the extraction of specific calculation results or desired synthesized values. For instance, an Algorithm Lens calculates an economic index that evaluates decentralization based on the percentage of token holders.

Relayer

Transfers data to other blockchains using threshold signing schemes like tECDSA. The destination is a data oracle pre-deployed on the target blockchain, which is synchronized in a key-value format at predefined intervals. ^[8]

Optimistic Data Oracles

Chainsight provides an on-chain data oracle for each supported blockchain. Networks have varying gas costs, ranging from expensive ones like Ethereum to more affordable sidechains, but minimizing on-chain data is crucial for any network to reduce gas costs. Additionally, most blockchains cannot access data outside their network, requiring an external data source, such as an Oracle contract, and someone to write data to that contract. Chainsight carries data within its network to other chains, eliminating single points of failure by combining distributed node’s derived keys into a threshold signature. ^[9]

Data Oracle contracts are pre-deployed on blockchains supported by Chainsight, allowing developers to access data cost-effectively without deploying the Oracle contract themselves.

To write data to the Data Oracle, developers first deploy a Relayer through the WebUI by specifying which data to write and at what time interval. The Relayer continually writes the specified data to the target chain's Data Oracle, using a one-to-one issued Data ID as the key. Any DApp on the target chain can retrieve on-chain insights by referencing the Data Oracle with the Data ID as an argument. ^[9]