Financial services
Budgets, maturity, and regulatory requirements differ significantly between processes. This situation favors companies that provide strong audit evidence and data to improve processes.
Processes fit for AI agents
- Tax and accounting reconciliation
- Credit document analysis
- Regulated customer service
- Accounts payable and receivable
- Fraud prevention and triage
Examples of units of work
- Invoice or entry reconciled
- Credit application analyzed with a decision trail
- Document checked
Recommended Skill Pack
Financial reconciliation and document analysis
Credit application received
Analyze the documents
Check documents and the customer record
- LGPD, art. 20
Final credit decision: a person with the authority decides
An identified person decides
Credit application analyzed with a decision trail
Billable unit
Evidence
When the other side has an agent
When a customer arrives through a comparison tool or an Open Finance agent, the bank's agent answers within its credit authority and leaves the trail the Central Bank expects.
Compliance
Compliance standard examples
- CMN Resolution 4.893/2021, as amended by CMN Resolution 5.274/2025 (National Monetary Council)
- Cybersecurity policy and requirements for contracting cloud processing and storage.
- Res. BCB 85/2021
- The regulator may veto cloud contracts. Where the agent runs is a regulatory matter, not only a technical one.
- LGPD, art. 20 (Brazil's data protection law)
- The data subject's right to review a decision made solely by automated processing.
Our read
Maximum opportunity for AI agents. Midsize banks, credit unions, card acquirers, and credit fintechs — where processes are document-intensive, regulations are well-defined, and processing costs are already monitored.