An AI agent that can move money is only useful if it can't move money you didn't authorise. OVAAL gives software agents scoped access to financial actions (bounded by policy, limits, approvals, audit, and revoke) so an agent operates inside an envelope you define, never with unrestricted authority over funds. Today the live capability is policy and automation; fully agent-initiated finance is in design with early partners.
OVAAL is a technology and orchestration platform, not a bank, EMI, or CASP. Regulated services are provided by appropriately authorised customers and third-party providers. Availability depends on jurisdiction, provider approval, and service scope.
We mark the line honestly. Policy and automation ship today; software agents initiating finance on their own is the roadmap we're building with design partners.
Define who or what may initiate a financial action, under which conditions, limits, and approvals. Rules preview before they run, every instruction is logged, and anything can be revoked. This governs human- and business-initiated flows now, and it's the same envelope an agent would operate inside. This page is built on the live policy & automation capability.
Software agents initiating financial actions on their own (within the same policy envelope) is in design with early partners. We're shaping the controls, the approval surfaces, and the audit model together, rather than claiming an autonomy we haven't shipped.
Scoped access means an agent can do exactly what you allowed, and nothing else. Five controls define the envelope.
What an agent is permitted to do: which actions, which accounts, which counterparties, under which conditions.
How much, how often, and within what windows. An agent can't exceed the bounds you set, even if instructed to.
Actions above a threshold route to a human or a second system for sign-off before anything executes.
Every agent instruction is logged with its context as one operational record, exportable for your compliance team.
Pull an agent's access at any time. Outstanding permissions stop applying immediately.
An agent acts within policy; it never gains custody of funds. Funds stay end-user-side, liability stays partner-side. See accounts & wallets.
Software agents may initiate financial actions only within explicitly defined policies, permissions, limits, and approval rules. Every instruction can be governed, logged, reviewed, and revoked.
OVAAL is the technology layer. Your license owns the regulated relationship. Authorised providers execute the regulated service. An agent acts inside your policy, it does not change who is responsible.
| Activity | Customer (you) | OVAAL | Authorised provider |
|---|---|---|---|
| Regulatory permissions | Primary for your business model | Technology role unless expressly authorised | Primary for the outsourced regulated service |
| Agent policy & limits | Policy owner | Policy engine and enforcement | Not applicable |
| Approval of agent actions | Approver, per your rules | Approval workflow | Execution eligibility |
| Custody of funds | Per your model; non-custodial by default | No, unless expressly authorised | Custody provider where applicable |
| Audit & evidence | Review and finance ops | Operational record and export | Source confirmations |
| Support & complaints | Primary | Technical escalation | Service-specific escalation |
Fully agent-initiated finance is in design with a small group of partners. Book an architecture review to walk through what's live, what's on the roadmap, and the controls we'd build around your use case. Prefer to read first? The integration brief is a one-page PDF covering modules, the responsibility model, and indicative commercials. Or browse all solutions.