Knowledge Graph

The heart of DeepBlock is its knowledge graph.

This is a purpose-built graph database that models blockchain ecosystems as an interconnected web of entities and actions.

In the knowledge graph:

• Nodes represent Entities

Key blockchain concepts become nodes in the graph. Examples of nodes include wallet addresses, smart contracts, tokens, blocks, transactions, and even higher-level entities like protocols or address clusters. Each node can have properties (e.g., an address node has a balance, a token node has a symbol and decimals, a transaction node has a timestamp and value, etc.)

• Edges represent Relationships

Connections between nodes are captured as edges. For example, if Address A sent 5 ETH to Address B in Transaction X, we create an edge like A --[transferred 5 ETH via Tx X]--> B. If a smart contract calls another contract, that call is an edge; if an address provides liquidity to a pool, that action is an edge linking the address and the pool. These relationships can be labeled and often carry attributes (amount, time, etc.).

Multi-Hop Connections

Because it’s a graph, you can naturally traverse multiple hops. For instance, from an address node, you can explore all outgoing transfers, then jump to the recipient addresses, and then see what those recipients did next, and so on. This makes it straightforward to answer questions like “What paths did funds take from point A, eventually reaching point C?”

Cross-Chain Integration

DeepBlock’s knowledge graph isn’t limited to a single network. We integrate data across supported networks. We handle this by either unifying common entities (for example, if the same wallet address (by private key) exists on multiple EVM chains, we can represent that as linked identities) or by connecting events through known cross-chain bridges (e.g., an event on Chain A that locks assets is linked to a mint event on Chain B that represents the same assets). The graph thus gives a holistic view of activity across chains. An analyst can query, for example, all transactions from an address across Ethereum, Polygon, and Arbitrum in one go.

Semantic Layer & Ontology

We define an ontology for on-chain data to ensure consistency. For instance, we categorize transactions by type (swap, transfer, stake, etc.), we classify addresses (EOA, contract, DeFi protocol, NFT collection, etc.), and we group related addresses into clusters when we detect on-chain heuristics that link them (for example, multiple addresses controlled by one entity might be clustered through analysis). All of this becomes part of the graph.

The knowledge graph approach is powerful because it mirrors how real blockchain interactions work, as a network of connections. It enables complex queries that would be very difficult with traditional relational databases or raw JSON APIs.

Queries like “find all addresses that interacted with a certain DeFi pool contract, then find other pools those addresses also used” become feasible and performant. Moreover, this graph is continually updated as new blocks come in, so it’s a living, breathing data structure representing the current state of the decentralized world.