01Industry model
It understands the things your business runs on.
A unit, a payment milestone, a snagging item, a policy, a claim status, a booking reference. Not words it has been told to recognise, but things with rules attached.
Every agent starts from a model of your industry: the things it runs on, the journeys customers take and the rules that apply. Built once for each industry, and better with every deployment.
Book a demoHow it's built
01Industry model
A unit, a payment milestone, a snagging item, a policy, a claim status, a booking reference. Not words it has been told to recognise, but things with rules attached.
02Workflows and journeys
An enquiry becomes a viewing, then a sale, then a handover. A claim moves from first notice to assessment to settlement. The agent knows which stage someone is at and what moves it forward.
03Policies as limits
Fees, entitlements, approval limits and fare rules are enforced by what the agent is able to do. It cannot offer terms outside them, because it has no way to construct them.
04Your live data
During the conversation, the agent looks up the real booking, unit or claim in the systems you already run, and answers from that.
05Arabic and English
Gulf, Egyptian, Levantine and Modern Standard Arabic, English, and many other languages your customers use, including customers who switch between them mid-sentence. It works out the language from the first few words and remembers each customer's preference.
06Tested before launch
Test sets drawn from real conversation patterns in your industry run before every release, scored against what a correct outcome means for your business.
07Governance and data
Claims outcomes, goodwill outside policy and anything involving safety stay with your team. Personal information is redacted before any data leaves your deployment, and every decision is recorded with its reasoning.
The same industry knowledge sits behind every agent. What changes is the job: winning the enquiry, answering the question, resolving the issue or starting the conversation first.
Answers enquiries at any hour, qualifies them against your criteria and books the next step with the right person.
Answers from your own systems and documents, the same way on every channel, every time.
Finishes the request, not just the conversation: the change made, the record updated, your team told.
Contacts customers first, for renewals, reminders and updates, only within the hours and lists you allow.
Every deployment in an industry adds workflows, edge cases and test conversations that the next one starts from. What carries across is industry knowledge. Your customers' data stays in your deployment and is never shared.
Questions
The industry model is a starting point, not a template you have to fit. We adapt it to how you operate during implementation, and your team reviews every workflow and rule before a customer hears it.
It says so, rather than guessing, and offers a person on your team. The conversation goes across with a written brief, so the customer doesn't start again.
Property, travel and hospitality, and insurance. Each has its own model of the things it runs on, its customer journeys and the rules that apply.
Every release runs against test sets drawn from real conversation patterns in your industry, scored against what a correct outcome means for you. Changes don't reach customers until they pass.
No. What carries across deployments is industry knowledge: workflows, edge cases and test patterns. Your customers' data stays in your deployment and is never shared.
Agents that hold up in real conversations don't come from a model release. They come from sustained research into how an industry works, how its customers behave and what the technology can reliably do. We do that work so your team doesn't have to, and every deployment starts from it.
Get started
One monthly figure per channel. Everything included.
We map your workflows, systems and rules, and design the agents around how your business already runs.
Live on your first channel within weeks, connected to the systems your teams already use.
Performance tracked against measures agreed with you before launch. Then extend, channel by channel.