Realtor Desk

AI in a real estate CRM: where it helps, and where it stops

Where AI helps in Realtor Desk: a lead score with its factors shown, an in-app assistant that drafts replies, and a plain list of what it does not do.

Realtor Desk uses AI in three specific places. It does not contact your leads, run a chatbot or forecast the market, and this page lists both sides so you can judge it.

Short answer

The AI ranks your leads, suggests a next step and drafts text for you to review. You stay the sender of every message. That is a smaller claim than most “AI CRM” pages make, and it is the one the product can back.

What the AI does

What it does not do

Why it is built this way

Under CASL the burden of proving consent is on the sender, so a system that messages leads on its own creates risk in your name. Keeping the human as the sender is a deliberate constraint, and it is why the assistant drafts rather than sends.

Showing the factors behind a score is the same instinct. A number you cannot inspect is a number you cannot argue with, and you know your leads better than a weighting does.

Questions

Related

See it on your own leads

14 days, then CAD $149 a month. A card is collected up front and nothing is charged before day 14.

A lead score with its working shown

Each lead gets a 0 to 100 score built from factors such as engagement, behaviour, budget match, timeline and qualification. The factors are listed beside the score, so you can see why a lead is ranked where it is and overrule it. It is a rule-based weighting, not a trained model.

A suggested next action

Beside the score the product suggests what to do next and when. It is a prompt for your attention: you decide, and you send.

An assistant on the contact you have open

Inside the app, the assistant can see the contact in front of you and draft a reply, summarise the thread, or talk through a next step, in English or French. It needs you signed in, and it works on one contact at a time.

Contact a lead on its own. Nothing is sent without you.

Run a chatbot on your website or answer leads out of hours.

Make or take calls. There is no voice agent.

Send scheduled email sequences. The builder exists, but scheduled sending is not switched on.

Predict prices, forecast the market or value a property.

Choose the best time to contact someone. It suggests; it does not optimise.

Measure its own accuracy. How well the score predicts outcomes has not been studied.

Is the lead score machine learning?

No. It is a rule-based weighting of factors such as engagement, behaviour, budget match, timeline and qualification, with the factors shown on the lead. That makes it explainable, and it also means nobody has tested how well it predicts a sale.

Does the AI reply to my leads for me?

No. The assistant drafts and summarises inside the app for the contact you have open. Nothing is sent to a lead until you send it, which also keeps you on the right side of CASL consent.

Is there an AI chatbot for my website?

Not today. The assistant is internal to the CRM and needs a signed-in user. It does not answer leads unattended.

Does the assistant work in French?

Yes. It can draft and summarise in English or French, and the language of a client's messages is recorded per contact.

Where does the AI process my data?

AI-assisted features send the content they work on to an outside AI provider, which may process it outside Canada. The database itself runs in Canada. Our PIPEDA page explains the difference between where records are stored and where each processing step happens.

See pricing or book a demo.