What is account-centric indexing?

Account-centric indexings aggregate all transaction history specific to each account, including transaction, trace, logs [1] etc, and keep them updated in real-time, on top of which UDFs execute to cache account-specific features allowing instant access through API calls.

The design brings in three main advantages that make Hemera Protocol an ideal public goods data infrastructure for web3:

1. Universal data interoperability: account-centric indexing tackles data fragmentation issues of multi-blockchain ecosystems. It achieves so first by aggregating account-specific blockchain data locally. Then it standardizes the downstream data pipelines, including RPC services, indexing, UDFs, APIs and AI & ML model interfaces, enabling developers to focus on building interesting data-driven features without heavy-lifting on underlying blockchain data infrastructure.

2. Highly extensible: Adding support for new blockchains becomes modularized and standardized. The UDFs design allows continued development and addition of new “features” and “scores”, and is independent from adding new blockchains support to the network.

3. Horizontally scalable: devices as small as personal cell phones hosting one account (effectively acting as a “personal data pack”), or as large as a single node hosting full Ethereum mainnet data (currently ~10TB in total as of end of 2023) can be seamlessly supported to join the Hemera network.

In the case of accessing multi-blockchain transaction history (e.g. Ethereum’s rollup ecosystem), account-centric indexing aggregates transaction histories from all blockchains under the same account. For example, to query a single account’s inscription token balances across all Ethereum Virtual Machine (EVM) chains over the last 24 hours, including mainnet, Layer 2’s, side chains, developers currently need to build in-house data pipelines to retrieve transactions from RPC services for each of the EVM chains, index and parse calldata from each transaction, and keep them updated in real-time. But with account-centric indexings, the transactions are aggregated, balances are pre-computed, and kept updated in real-time through built-in inscription UDFs, developers simply need to call the APIs provided by Hemera.