Related ToolsClaudeClaude Code

Anthropic Ships 10 Finance Agent Templates for Banking, Credit, and Investment Teams

Increasing bar chart with node lines and arrow tracing the bars
Image: Anthropic

100% of Walleye Capital's 400-person hedge fund uses Claude Code. That adoption number helps explain why Anthropic is now shipping purpose-built agent templates for the sector rather than leaving firms to build from scratch.

Anthropic announced in a blog post ten ready-to-run agent templates for financial services and insurance work. The templates split across three categories: research (pitch builder, meeting preparer, earnings reviewer, financial model builder), credit and risk (market researcher, KYC screener), and operations (valuation reviewer, general ledger reconciler, month-end closer, statement auditor). Each one is wired up to run inside Claude's existing Cowork and Code products, or via Claude Managed Agents on the API platform.

What the Templates Actually Do

The most concrete result in the announcement comes from FIS, a financial technology company, which says AML (anti-money laundering) investigations that previously took days now complete in minutes using Claude. Mizuho, the Japanese bank, described their analysts' prep work as having been converted into idea time. These aren't vague efficiency claims - they point to specific, repetitive workflows where the agent handles the document pulling and initial drafting while a human reviews.

The data integrations are where this gets practically useful. Anthropic built connectors to Dun & Bradstreet, Moody's, Verisk, SS&C IntraLinks, and several other financial data providers via MCP (a protocol that lets AI models plug into external data sources). The Moody's connector alone covers over 600 million public and private companies. There's also a Microsoft 365 add-in that works across Excel, PowerPoint, and Word today, with Outlook support coming soon - meaning analysts can carry context across tools without re-explaining what they're working on each time.

On benchmarks: Claude Opus 4.7 scores 64.37% on Vals AI's Finance Agent benchmark, which evaluates how well AI agents handle realistic financial tasks. That's a specific, verifiable number, which is more useful than the usual qualitative claims.

Who's Already Using This

The customer list is not a collection of pilot programs. Citadel, BNY, Carlyle, Mizuho, Travelers, and Walleye Capital are named as production users. These are large institutions with compliance obligations and risk management requirements that make them conservative adopters - their presence suggests the governance controls (audit trails, human approval steps, governed data access) are solid enough for regulated environments.

The templates are available now on paid Claude plans. Managed Agents on the API platform are in public beta.