A confidential AI workspace for mortgage brokers

A mortgage application is a person income, debts and ID in one file, and pasting it into a public chatbot sends all of it out of your business. That payslip, that passport scan, that bank statement summary, the moment it lands in a chat box hosted by someone else, it has left the building and you cannot pull it back.

You do not have to choose between a capable AI and keeping client files on your side. You can run a strong model on hardware you control, so a question about a named applicant stays with you. This page shows how brokers do exactly that.

Why a public cloud chatbot clashes with brokers handling applicant income and ID

A public chatbot is a service in someone else cloud, usually in the US. When an adviser pastes a fact-find or a salary figure into it, that text travels to a server you do not own and cannot inspect. For most office work, fine. For a file that names a real applicant and their finances, that is sensitive client data going somewhere you have no view into.

Banning AI outright does not solve it. Your staff already use these tools, on their phones, in a browser tab, to draft an email or summarise a statement faster. Tell people they cannot, and they will quietly carry on. The fix is not a ban. It is giving them an AI that is just as quick and lives on your side.

Run the model in-house

With kral, the platform runs on your own server. You can add a local model on your own hardware, so a prompt about a named applicant goes to your machine and stops there, with no external API in the path. The income, the ID, the debts, none of it leaves the room. Most brokers mix the two: a cloud model for general work like rewriting a marketing blurb, and a local model for the sensitive cases where an applicant name and numbers are in the text. You decide which work goes where.

A full workspace, not a chat box

This is more than a box you type into. Your team can build their own assistants in minutes with no code. One adviser sets up an assistant that drafts client-facing offer summaries from the key figures, so every client gets a clear, consistent write-up. Another builds an assistant that turns a fact-find into a clean file note, ready for the case record. Save those as reusable routines and nobody rebuilds the same setup twice. Drop in a document and ask questions about it. Pull a current, cited answer from the web when you need a rate or a rule checked. Switch between the leading models in one click, depending on the job in front of you.

Connect your own systems

kral supports MCP, the open standard for connecting tools and data to an AI. That means the assistant can work with your own templates and internal knowledge through a connector you control, instead of guessing from the open web. Ask it to produce an offer summary in your house format, or answer from your own criteria notes, and it draws on your material. Your systems stay yours, and you decide what the assistant can reach.

You run it and you see everything

You manage who is in and which models each person can use. Set a spending limit per adviser so costs never run away. Watch real usage on a dashboard, so you know what is being used and by whom. Staff sign in with single sign-on, so there is one account to manage. It installs on Windows Server behind IIS, sits inside your network behind your firewall, and wears your own branding. If you want the wider picture of company-wide AI you host yourself, the same approach scales across the whole firm.

We help you put it in place

You do not set this up alone. We install kral with you, connect it to your systems through MCP, and advise on rolling AI out across the team without the data leaving your side. Implementation consulting is part of what we offer, so the platform is working the way your office actually runs, not left as a manual.

Capable AI and client confidentiality are not at odds. Keep the income, the ID and the fact-find on your server, give your advisers a fast tool they will actually use, and stay in control of every part of it.

Book a demo

Open the app