{"slug":"protect-crypto-assets-with-governed-ai","url":"https://findausecase.com/use-cases/protect-crypto-assets-with-governed-ai","title":"Protect Crypto Assets with Governed AI","description":"Elliptic, a blockchain analytics and digital asset compliance company, built a copilot on Databricks Agent Bricks to reduce compliance analyst workload by describing wallet activity, highlighting exposure paths, and producing neutral, report-ready risk narratives for financial institutions, crypto businesses and government agencies. MLflow tracing captures prompts, intermediate steps and final narratives for every interaction, and automated evaluation pipelines score responses for correctness, relevance and safety using LLMs as judges. Domain-specific safety guidelines distinguish between describing financial-crime typologies and inadvertently mentioning harmful activity, while custom safety scorers distinguish between explaining money-laundering patterns and promoting harmful activities. Elliptic reports compliance teams resolving high-risk screenings up to 2x faster and 40% less time needed to complete a Suspicious Activity Report.","company":"Elliptic","industry":"Technology & Software","aiCapabilities":["Generative AI"],"businessFunctions":["Legal & Compliance"],"technology":["Agent Bricks","MLflow","Databricks Foundational Models API","Model Serving"],"deployment":"Public Cloud","problemStatement":"Rapid growth across crypto markets, rising transaction volumes and expanding regulation place additional pressure on Elliptic's clients: compliance teams at exchanges, banks, payment providers and institutional custody platforms. Compliance teams needed strong assurances that Elliptic's copilot would not hallucinate, omit key context, or rely on opaque reasoning, particularly when its outputs might feed into formal investigations and regulatory reports.","solutionApproach":"Building on its existing Data & Intelligence Platform, Elliptic leveraged Agent Bricks to build Elliptic's copilot, designed to reduce compliance analyst workload by describing wallet activity, highlighting exposure paths and producing neutral, report-ready risk narratives, all aligned to client risk appetites. MLflow tracing captures prompts, intermediate steps and final narratives for each interaction, and automated evaluation pipelines score responses for correctness, relevance and safety using LLMs as judges, with custom safety scorers distinguishing between explaining money laundering patterns and promoting harmful activities.","businessValue":"Compliance teams are resolving high-risk screenings up to 2x faster, without sacrificing accuracy or documentation quality, and Elliptic reports 40% less time needed to complete a Suspicious Activity Report.","evidence":{"band":"high"},"sourceUrl":"https://www.databricks.com/customers/elliptic","dates":{"publishedAt":"2026-09-05T09:05:03.737Z","publishedAtSource":"pipeline","updatedAt":"2026-09-05T09:05:03.737Z"},"license":"Open for reading and citing with a link to https://findausecase.com/use-cases/protect-crypto-assets-with-governed-ai. Bulk republication requires permission."}