Digital asset compliance
Chainward
The alert with its reason attached
- Sector
- Digital asset compliance
- Location
- Dubai, UAE
- Year
- 2025
- Kickoff to production
- 15 weeks
Wallet screening and case management for an exchange whose compliance team was reading block explorers by hand.
Every withdrawal review meant an analyst pasting addresses into three tools and forming an opinion. Cases lived in a spreadsheet, thresholds were personal habits, and a regulator's question about a decision from March had no answer.
- 01
Exposure, measured
Direct and indirect exposure is computed against labelled clusters with the hop distance stated, so an analyst sees magnitude instead of a colour.
- 02
Cases that write themselves
The graph, the risk factors and the transactions that triggered the review are captured into the case at the moment it opens.
- 03
Thresholds owned by policy
Escalation rules are versioned and dated, and any past decision can be replayed against the policy that was live at the time.
Review time per withdrawal fell to a few minutes, and the team's case files became the primary evidence in its next supervisory review.
What changed
- Withdrawal review
- Minutes
- On every case
- Stored graph
- Decision reproducibility
- Full replay
Stack
- Python
- Postgres
- Neo4j
- Next.js
- Airflow
Disciplines applied
- AI solutions
- Data & analytics
