§1 — The brief.
The brief.
Canton Network is the institutional finance chain operated by Digital Asset. The protocol's contract layer runs on Daml — a language and runtime designed for multi-party financial workflows where privacy boundaries between parties matter as much as the contract logic itself. That privacy model has a consequence for indexing: a generic blockchain indexer that walks all transactions and builds a public state view is wrong by construction. Canton indexers must respect the privacy contract that the underlying chain enforces.
The engagement: build indexer infrastructure for Canton — Daml-aware, privacy-respecting, with a query API suitable for explorer-grade access patterns and institutional tooling.
The design came out of hands-on Canton work — Zpoken deployed an RWA tokenization project on Canton and worked through the Ledger API, PQS, and the sub-transaction privacy model directly.
§2 — Shipped and in flight.
Shipped and in flight.
[01]
Canton Network Indexer — in active development: three tiers (public Scan data · party-scoped contracts via Ledger API/PQS · cross-party aggregation with RBAC)
─ IN DEVELOPMENT
[02]
Working PoC — schema, GraphQL API, party isolation, lifecycle queries, JWT party-scoped auth
─ PUBLIC REPO
[03]
RWA tokenization deployment on Canton — the hands-on work the indexer design came from
─ DEPLOYED
The indexer is being built now. The PoC repository is public; the engagement work itself is private.
§3 — Engineering decisions.
Engineering decisions.
Privacy-first storage.
Canton's contract layer is built around party visibility — not every participant sees every event. A generic EVM-style indexer built around "all transactions are public, fold them into a state tree" produces the wrong artifact for this chain. The indexer here is built directly against Daml's semantics — contract templates, party visibility, exercise events — so the indexed state respects the privacy model rather than collapsing it. The data is queryable, but only along the lines Canton itself permits. Every query is filtered by party claims at the SQL level — signatories and observers are first-class columns, and the indexer sees only what the participant's parties can see.
Daml-aware, not chain-generic.
The pipeline is built against Daml as a first-class input, not retrofitted from a generic on-chain indexer. Contract templates, choices, and exercise events are modeled directly — the indexed state is a faithful Daml projection, not a flattened transaction log. Contracts are indexed on Daml's own lifecycle — create → active → archive, eUTXO-style, never mutated — with template-specific views generated from compiled DAR files.
Query API shaped to institutional tooling.
Explorer-grade access patterns for institutional finance look different from public-chain block explorers. The API ships the queries that institutional ops, audit, and analytics actually run, not generic on-chain primitives. This trades flexibility for fit-to-use. GraphQL primary (its type system maps to Daml templates), REST for lookups, SQL for power users. Ingestion builds on Digital Asset's PQS rather than a custom gRPC client — the post-processor extends it instead of rewriting it.
§4 — Where it stands.
Where it stands.
Active, 2026. The indexer is in development — the public PoC describes the architecture; the build itself stays off public GitHub, as Canton work does. No public product surface to link yet.
§5 — Tech stack.
Tech stack.
~/zpoken/work/canton/stack
Daml · Canton Network · indexer pipeline · query API · operational monitoring
§6 — Public surface.
Public surface.
§7 — Founders.
Founders.
Engagement led by Mike Yezhov and Anton Yezhov.