How they use Tavily

BMO uses Tavily to retrieve public records for anti-money-laundering screening. Analysts search an entity name and receive ranked URLs with relevant passages extracted. Tavily can retrieve full-page content for deeper reviews. BMO’s own models assess possible risk, and investigators make the final decision. BMO technology and Tavily jointly built the retrieval layer. BMO reports that Tavily helps make retrieval results consistent across investigator actions. Tavily says the secure web retrieval layer helps analysts complete more reviews and determine risk faster. Tavily says Ashutosh Sinha turned to it because keeping up with AML research demand required either adding analysts or automation.

Tavily’s September 2026 account describes AML as BMO’s first production use case. Search volume grew from roughly 40 calls in the first month to more than 129,000 in July 2026, with more than 97,000 calls in each of the preceding six months. The integration is API-only. Tavily fit an architecture pattern BMO’s security team had already approved. BMO’s security review only had to happen once, and new use cases can build on the same approved foundation. The account reports expanding Tavily adoption as other BMO teams build and deploy agents.

Tavily’s October 2026 account reports hundreds of thousands of monthly queries in BMO’s AML workflows. It describes Tavily supporting an AML risk-scoring model and providing a common search and web retrieval layer for agents across the bank.