Under the partnership, Regula and Bernini.ai have integrated their technologies so that AI agents running on Bernini.ai's DAFNE platform can call Regula Document Reader SDK and Regula Face SDK whenever a customer's identity needs to be established or reconfirmed. The integration covers the full customer relationship, from onboarding to servicing existing accounts and processing transactions that require a higher level of assurance.
Regula develops identity verification solutions, while Bernini.ai is a Latin America-based provider of AI agents and agent management technology. The joint offering targets regulated and transaction-intensive sectors, including banking, insurance, telecommunications, and retail, where customer identity, business rules, and AI agent actions need to operate within a single workflow.
How verification fits into agent conversations
A DAFNE-powered agent can handle a range of customer requests, from opening a new account to assisting with an existing one. When the business needs to confirm who it is dealing with, DAFNE triggers Regula's document and biometric checks. It then applies the organisation's rules to the verification result to decide how the interaction should proceed.
Verification is not limited to a single point in the customer lifecycle. During onboarding, the check establishes a new customer's identity. In later interactions, DAFNE can request a fresh check when a customer asks about an existing account or initiates an action that calls for stronger assurance. For example, a customer verified at onboarding may be asked to verify again before a later request involving that account proceeds.
In addition, Regula's document verification draws on the company's decades of forensic experience and extends beyond data extraction from an ID. Document Reader SDK examines all document data fields and security features, then cross-checks the information to detect inconsistencies that may indicate fraud. Verification data can also be processed and re-verified on the server, adding a further check within a trusted environment.
Face SDK links the document to its holder through face matching and applies liveness checks to confirm that a real person is present during verification.
Governing agent actions
Once Regula returns a result, DAFNE uses it as an input to its orchestration and decisioning logic, determining whether the workflow continues and which actions an agent may take at that stage. Instead of a single general-purpose agent, DAFNE deploys specialised agents for different tasks. A shared control layer coordinates their interactions and applies the organisation's authentication and authorisation policies, business rules, security controls, data-protection requirements, and audit logging.
This structure allows organisations to define what each agent can access and do, manage handoffs between agents, and keep a record of their actions. DAFNE supports interactions across channels, including web and mobile, with Regula's verification available wherever the workflow requires it.
Identity as an ongoing control
The partnership reflects a shift in how identity is treated in agent-driven services. Henry Patishman, Executive Vice President of Identity Verification Solutions at Regula, described AI agents as an emerging channel through which customers open accounts, access information, and request transactions, noting that organisations need evidence of who is in the conversation at critical moments. Tomas Alburquenque, CEO and Founder of Bernini.ai, said identity should not be verified only once at onboarding, as risk changes across the customer journey and some actions demand stronger assurance. The company aims to extend this model to additional industries and markets.
For regulated businesses deploying AI agents in customer-facing roles, the integration links verification outcomes directly to what an agent is permitted to do, placing identity checks inside the workflow instead of before it.
