Our products.
Every product opens its evidence: what the Co-pilot finds, the Smart Graph shows, and the Risk Ledger scores all cite the intelligence underneath.
Investigator's Co-pilot
One question box, in plain language. An AI investigator drives the platform's full forensic toolkit and returns a readable, cited answer.
Explore →
Smart Graph
Follow the money visually with an interactive graph of addresses, transactions and entities, with every element tied to the evidence beneath it.
Explore →
Risk Ledger
Mass screening for compliance teams: every score opens up to "Why this score", with evidence links, a defensible audit trail, and an API to bring it into your own tools.
Explore →
Ask it like a colleague. It investigates like a team.
One question box, in plain language. Behind it, an AI investigator drives the platform's full forensic toolkit — from tracing, clustering and attribution to intelligence lookups and risk checks — and comes back with a readable answer that cites its evidence.
Plain language in
"Where did this victim's funds go?" is a complete query. No graph-tool expertise; no query language; usable on your first crypto case.
A full toolkit
The Co-pilot chooses and drives the right forensic tools for the question, with the same tracing, clustering and intelligence machinery an expert would use, in the right order, automatically.
Cited answers out
Every claim in the answer carries its source: the on-chain structure, the sanctions entry, the intelligence record. Follow any citation to the evidence itself.
Follow the money. Literally.
The Smart Graph is the investigation, drawn: addresses, transactions and entities as an interactive network. Expand a cluster, follow a hop, and watch flagged funds resolve, with every node and edge tied to the evidence beneath it.
Explore, don't decode
Clusters, exchanges and flagged actors are labelled in place. Red means illicit, amber means watchlist, green means verified, with evidence colours consistent across the platform.
Built on attribution
Nodes aren't just addresses: ownership clustering and named-entity attribution, group them into the actors behind them, with clear provenance.
From picture to proof
Any path you find on the graph can go straight into a cited report. The visualisation and the corresponding blockchain intelligence are presented together.
Screening at scale. Evidence per verdict.
Mass screening for compliance teams: score addresses and counterparties, decide, and defend the decision — every verdict keeps the evidence and derivation that produced it.
Click any score. See the whole case.
Every risk score opens into its full derivation: each contributing factor with the score it alone would give, the factor that set the final number marked, and evidence links straight to the source, with a sanctions entry, a published attribution, and an intelligence record.
Verdicts are frozen at screening time with the derivation stored, so what an officer decided and why is an audit record, rather than a mutable opinion.
Named-entity attribution. Scores rest on named entities and never anonymous blanket labels, with source linked and confidence stated.
Decisions, audited. Clear, escalate or whitelist, recorded with who, when and why. Whitelisting requires a written reason.
Re-screening that respects you. New intelligence flags what changed, and never silently alters an officer's decision.
Recommendation escalate under Standard AML screening v1 — auto-clear below 61 · escalate at 81+
Integrate the intelligence, not another dashboard.
For screening at scale, the API is the product: bring Sapphire Trace's verdicts and evidence directly into your own onboarding flows, case-management systems and compliance tooling.
Score + evidence, per call
Every response carries the verdict and its "Why this score" derivation, including factors, sources, confidence. The glass box travels with the API.
Fits the stack you have
Screen at customer onboarding, on transaction events, or in batch over a book of counterparties, and route verdicts into the tools your team already works in.
AI-native by MCP
The same capabilities are exposed over the Model Context Protocol, so your own AI agents and copilots can drive Sapphire Trace directly.
See it on one of your cases.
The fastest way to evaluate Sapphire Trace is to bring a real investigation to a demo.