← All work
Web3
Blockchain analyzers for compliance and research
Pipelines that turn raw blocks into address graphs, labels and risk scores across several chains.
- Client
- Analytics products under NDA
- Field
- Web3
- Our part
- Full cycle: ingestion, graph, scoring, API and UI
- Stack
- RustPythonClickHouseNeo4jKafkaTypeScript
The problem
Raw chains are just transfers between hashes. Compliance teams and researchers need entities: which addresses belong together, which touched a sanctioned service, how funds moved through mixers and bridges, and all of that within seconds of a new block.
What we built
- Multi-chain ingestion with reorg handling, so the graph never lies about a block that was later dropped
- Address clustering with heuristics and labelled seeds, kept explainable so an analyst can see why two addresses were joined
- Flow tracing across mixers, bridges and exchanges, with risk scoring exposed over an API
- An explorer UI for analysts and alerting for compliance teams
What it changed
- Analysts get an entity graph with reasons, not a black-box score
- The ingestion design now backs our indexers for client wallets and exchanges