Under the channels your clients see sits one design: a governed core your brokerage runs in isolation, an industry module that carries your market's rules, and tools and AI agents that read whichever module is active. It all renders under your brand and pushes structured events into the stack you already run.
White-label here is not an "upload your logo" setting on a TomoBroker-branded page. The identity is applied end to end, from the domain to the colours on the page, and shared from one place rather than re-entered per tool. Here is what is live, said plainly.
A brand kit is your brokerage identity: name, logo, colour palette, a curated font pairing, and optional partner co-branding. Every client-facing touchpoint resolves its look from a kit today, so one change re-skins every touchpoint at once. Every brokerage gets a default kit automatically.
Every brokerage gets its own {brokerage}.tomobroker.com the moment it is provisioned, with reserved system names protected so no brokerage can claim them. Public funnels and embeds load under that branded subdomain with no login step.
The custom-domain infrastructure is shipped: a brokerage points its own domain at TomoBroker, ownership is verified by a DNS record, and the SSL certificate is issued and renewed automatically. The channel served on a custom domain today is the Client Concierge hosted chat page. Serving Scout funnels and Spark widgets on a custom domain is the next step on the same infrastructure; today those serve on the brokerage subdomain.
Three ideas are kept separate: the industry (the broker's domain), the country (the regulatory geography), and your brokerage. Each industry-and-country combination is served by a self-contained module carrying that market's data, matching logic, calculators, compliance rules, and terminology. The tools and AI agents know nothing about any one domain; they read whichever module is active.
So entering a new industry or country is one new module against a stable contract, not a fork of the app. The matching engine, the qualification funnel, and the calculators get parameterised, not rewritten.
Exactly one market is built and live today: Australian mortgage broking. It is the reference implementation: the calculators behind Spark, the fact-find and live-lender matching behind Scout, and the AU-mortgage compliance rules the AI agents follow.
Other industries (insurance, asset finance, commercial lending) and other countries (US, NZ, UK) are the seam and the direction, not live markets. The contracts, the registry, and the industry-agnostic tools are proven against one real market, so the second one plugs into an existing shape.
TomoBroker doesn't ask you to tear out the systems you already run. Your CRM, aggregation software, lender panel, and lodgement stack stay in place. TomoBroker sits in front of them as the capture-and-qualify front door and pushes signed, versioned events into that existing stack. Adopting it is additive.
The integration is outbound webhooks: point TomoBroker at an HTTPS endpoint and it delivers a JSON event the instant something happens. Each is signed with the Standard Webhooks scheme so the receiver can verify authenticity, retried with backoff on failure, blocked from internal addresses, replayable, and metered per plan. Every event carries provenance: which tool or AI agent produced it, and the assigned broker where relevant.
The v1 event catalogue
Nine events, one signed envelope
scout.submission.createdenquiry.createdenquiry.allocatedenquiry.lifecycle_transitionedscheduling.booking.createdclient.createdclient.company.createdai_agent.session.startedai_agent.handoff_requestedNew enquiries land in the brokerage's CRM in real time; a stage change triggers its own automation, with no polling and no re-keying between systems.
Broker-domain work lives in reusable features, not baked into one screen or prompt. A tool packages them for you to assemble; an AI agent composes them into a role. Both call the same vetted engines the dashboard uses, so an agent answer and a broker answer agree.
What the client touches. Tools are packages you assemble; AI agents are workers that compose the same jobs and talk to the client.
Reusable broker-domain engines. Calculations and matching are deterministic, so the same inputs always give the same result, whether a tool or an agent calls them.
Every run happens inside one brokerage, bound to its active industry module, under its permissions and an append-only audit trail. Nothing sits outside that model.
A model chooses which feature to call, never what number to say. Calculations and matching come from the engine, so they cannot be invented.
Every run is scoped to one brokerage under its permissions, so one brokerage can never reach another's data.
Each run records what was asked, which features ran, and the facts relied on, so "what did the client see then?" can be answered.
Spark and Scout on your pages, the Client Concierge answering 24/7, every lead one enquiry flowing into your own stack. Start free, or book a demo to see the whole picture.