Most businesses were not built for software to act on them.
They were built around human availability — a phone call, an email, a contact form, a booking request, a message in an inbox.
A person on the other side, reading, interpreting, replying, confirming, quoting, taking payment, and moving the work forward.
That is still how millions of businesses operate. A restaurant takes orders by phone. A salon confirms appointments by message. A clinic answers availability by email. A repair shop gives quotes manually. A local service business turns customer intent into action one request at a time.
The web made these businesses visible, but not truly actionable. A website can tell a customer what a business does. A search result can help someone find it. A form can collect a request. A marketplace can make a narrow category transactional.
But most businesses still do not expose their real operating rules to software: what they sell, what is available right now, where they deliver, what it costs, which options are allowed, what information is required, whether payment can be authorized, what happens after a request is submitted, and when the business accepts, rejects, modifies, or fulfills it.
Today, AI agents cannot reliably answer those questions for most small and medium-sized businesses. Not because the agents lack intelligence, but because the business has no structured interface for them to use.
The information is scattered across websites, PDFs, menus, booking tools, delivery platforms, emails, social pages, and human memory. The transaction is trapped behind channels designed for people: phone calls, inboxes, forms, and manual back-and-forth.
That is the bottleneck.
AI agents are becoming a new interface between people and the world. Customers will increasingly ask agents to do things for them: order dinner, book an appointment, compare options, request a quote, schedule a service, authorize payment, and track what happens next.
But for agents to act, businesses need to become legible to them. Not just searchable. Not just listed. Not just “online.”
Actionable.
A business that is ready for AI should be able to expose, in a trusted and structured way, what it offers, what its rules are, what actions can be taken, what requires confirmation, and how a transaction should safely move from intent to completion.
That is what Riviera is building.
Riviera gives businesses an agent-ready storefront: a machine-readable operating layer that lets AI agents discover the business, understand its products and services, create accurate quotes, check constraints, request confirmation, authorize payment, submit orders, and retrieve status — while keeping the business’s own rules intact.
For the customer, the experience becomes simple: “Order dinner for four, keep it vegetarian-friendly, under £50, and have it arrive before 8.” “Find a salon near me with availability tomorrow afternoon and book the first suitable slot.” “Get a quote to fix my boiler and only confirm once I approve the price.” “Order my usual from the local restaurant, but make sure they still deliver to my new address.”
The agent handles the details. The business receives a structured request. The customer stays in control. The business stays in control.
For the business, Riviera becomes the infrastructure layer between old-world operations and the AI-native economy. Instead of rebuilding for every AI platform, each business gets one endpoint that agents can use — a place where menus, services, availability, rules, pricing, fulfillment, payment, and status can be exposed safely.
An AI storefront is not a chatbot. It is not another marketplace. It is not a new inbox to manage.
It is an interface for action.
A restaurant can receive AI-generated orders with the right menu, delivery zone, modifiers, price, customer confirmation, and payment authorization. A salon can expose services, staff, availability, booking rules, deposits, and cancellation policies. A clinic can let agents check appointment types, intake requirements, and available slots without exposing sensitive workflows. A repair business can collect structured job details, create estimates, authorize deposits, and route requests to the right person. A local operator can become instantly understandable and transactable by software without hiring an AI team or integrating separately with every assistant.
This is how millions of businesses move from being passively online to actively programmable.
The future should not belong only to marketplaces, aggregators, and the largest platforms with the resources to build agent infrastructure themselves. Independent businesses should be reachable by AI on their own terms, through their own domain, with their own rules, prices, and customer relationships preserved.
We’re starting with one restaurant and one pizza order because the future has to work in the real world.
If an AI agent can order from a local restaurant — read the menu, follow the delivery rules, create the right quote, ask for confirmation, authorize payment, send the order to the merchant, and report the final status — then the same pattern can extend far beyond restaurants: to salons, clinics, tradespeople, tutors, hotels, repair shops, local services, independent businesses everywhere.
The web made businesses visible.
Riviera makes them actionable.