How does a real estate agency train the intusell AI? A step-by-step guide

A real estate agency's inbox is always full of the same questions: "Is this flat still for sale?", "How many square meters, how many bedrooms?", "Is it mortgage-eligible, what's the title-deed status?", "Will you show it on the weekend?". Property portals, Instagram posts, and the WhatsApp line — it doesn't matter; the same twenty questions repeat a hundred times a day. While the agent is at a viewing, at 11 PM a buyer writes "are you free tomorrow?" and waits until morning for an answer. Every listing question answered late goes to another agency.

intusell takes over this load. But to do that, it first needs to learn your agency, your portfolio, and your operating routine. This article explains how a real estate agency trains the intusell AI from scratch, step by step. This is the real estate chapter of our sector-based "how to train your AI" series, and it is the series' main article.

Quick answer

Why does training matter?

intusell is not an off-the-shelf chatbot; it is a fully autonomous AI salesperson that behaves like an experienced office assistant and lead handler. A good assistant does not start on day one without knowing the portfolio either: they need to know which listings are active, which areas you work in, and which questions to leave to the agent. Training ensures the AI does two things correctly: give the right information (listing details, area information, working hours) and stop at the right boundary (not invent a price/title-deed/mortgage matter it is unsure of, and leave negotiation to the agent).

An untrained assistant either gives overly generic answers or guesses on a topic it doesn't know. In real estate, guessing is expensive: a wrong square-meter figure or a groundless sentence like "yes, it's mortgage-eligible" both brings a buyer to a wasted viewing and shakes the agency's credibility. That's why the most critical part of training is teaching the AI "what it knows" and "what it must hand off to the agent when it doesn't." A well-trained assistant stops when it doesn't know.

Who is it for?

This guide is for real estate teams whose inbox is dominated by listing questions and viewing requests, and who answer the same questions over and over again:

• Real estate agencies selling residential and commercial property
• Solo or small-team real estate agents
• Rental-focused offices with high listing traffic
• Teams receiving dozens of leads a day from portals and Instagram
• Agents handling many simultaneous buyer prospects in project/investment sales
• Agencies selling to foreign buyers from abroad and communicating in multiple languages

You don't need a technical team. All of the steps below are completed from sections in the panel, without writing code.

The structure of training: 5 core steps + 2 continuous steps

Splitting the training into two groups makes your job easier. The first 5 core steps get the assistant up and running; these steps are the one-time part of the setup. The 2 continuous steps that follow (label review and opening channels) are the ongoing part that you sharpen as you use the assistant. You'll find all of them in order below.

1. Upload the portfolio knowledge base

The foundation of training is the knowledge base (RAG). In the panel's Knowledge Base section, you upload all of your agency's textual information. Each time the AI generates a response, it automatically searches these sources and uses only the information written there. In real estate, this means your portfolio, directly.

Supported source types:

Every file you upload is automatically chunked and made searchable with pgvector. When you upload an "active portfolio.xlsx", the AI finds the right rows and sorts the suitable listings when a customer asks "what do you have that's 3-bedroom and under 2 million?".

What a real estate agency must definitely upload: the active listing list (number of bedrooms, square meters, floor, price, for-sale/for-rent status), area and transport information, project details and floor plans, payment/mortgage/title-deed process information, working hours, and the most frequently asked questions. An important limit: the knowledge base is a live source that must be kept up to date. If a sold flat hasn't been removed from the portfolio, the AI still thinks it's active; so keeping listing statuses current is the foundation of the correct answer.

2. Set up property viewing appointments and working hours

The knowledge base explains "what you have"; the appointment system, in turn, lets the assistant give the customer an actual viewing time. In the panel you define three things:

Once these three are defined, the AI no longer gives times out of thin air; it suggests viewings based on your working hours and full slots. Two appointments at the same time (double-booking) are prevented by the system — so you don't get the accident of bringing two buyers to the same flat at the same time. When an appointment is created, if Google Calendar is connected, the appointment is also copied one-way to the agent's calendar. The customer can reschedule or cancel their own appointment via the /manage-appointment/{token} link sent to them; this reduces last-minute wasted trips.

We've gathered all the details of the appointment engine — durations, travel buffers, conflict prevention, reminders, and calendar sync — in a separate article: portfolio matching and property viewing appointment automation. You can also see the portfolio and appointment capabilities on the solutions page and the real estate solution page.

3. Set the persona and tone

You taught the AI what to say in the first two steps; now you'll define how it will say it. There are two settings in the panel:

• Assistant name (ai_persona_name): The name it will introduce itself with to the customer. Most offices choose a reassuring, friendly name consistent with their team.
• Tone (ai_tone): For real estate, the typical tone is warm, solution-oriented, and fast; an advisor's language that tries to understand the buyer's intent and moves them toward a viewing without pressure. You can make the tone more corporate or more casual to match your own office culture.

A multilingual note: The AI responds in whatever language the customer writes in. For foreign buyers writing from abroad in English, Arabic, or Russian, you do nothing extra; the assistant automatically detects the language and responds in that language while preserving your tone — a lead that would otherwise slip away becomes a booked viewing.

4. Define the response rules and limits

The persona defines "who it is," and the response rules define "how it will behave." In real estate, this step does two jobs at once: conveying your sales discipline and drawing where the AI should stop and hand off to the agent. With the AI Manager Chat, you add rules by writing in plain language, just as you would instruct a teammate.

Typical rules for real estate agencies:

• "Before a viewing, learn the customer's budget, bedroom needs, and area, and suggest 2-3 suitable listings."
• "Don't close the negotiation and final price yourself; direct them to a meeting/viewing with the agent for the listing they're interested in."
• "Don't state square meters, title-deed status, or mortgage eligibility you're unsure of; say you'll ask the agent and hand off."
• "Always pass negotiation with the owner, deposits, and contract processes to the agent."

The most common mistake in real estate is the AI stating information it doesn't actually have (for example an out-of-date price or a promise like "definitely mortgage-eligible") as if it were true. In intusell, two layers work against this. The primary control is this: by system instruction, the AI is constrained from the start to use only the information in the knowledge base and to hand off when unsure. The second layer is a safety layer that scans responses and flags risky phrases; it runs in observation (shadow) mode by default, so it does not hard-block the response — it detects and flags. The real assurance is that the AI is constrained not to produce inflated promises in the first place.

If you want to compare two different approaches, you can use the A/B test feature: for example, you can put a more informative tone next to a closing that more directly calls for a viewing at 50%-50% traffic and measure which one converts to more appointments.

5. Teach from past conversations

This is the step that moves training from "good" to "specific to your agency." In the panel, you upload audio recordings of your past sales conversations (MP3, MP4, WAV, M4A) and mark each one as Won or Lost — for example, whether the customer came to the viewing and made an offer, or was lost.

The system uses these recordings in two ways:

This way, the AI learns how your best agent uncovered the buyer's budget, moved someone who said "let me think about it a bit" toward a viewing, and handled a price objection. The KVKK/GDPR side is protected: PII (personal data) masking and explicit consent are applied to the uploaded recordings.

This step is not mandatory, but don't skip it. The knowledge base teaches the AI "what" it knows; past conversations teach "how you sell." When the two are combined, the assistant truly resembles an experienced real estate agent.

6. Correct with label review (continuous step)

The first five steps get the AI up and running. From here on is the continuous part that perfects it over time. Every AI response lands in a label review queue. Here you can do three things: approve, reject, or correct.

The AI learns from these corrections. Say it explained an area's transport links incompletely or stated a project's delivery date wrong; you correct it, and in similar cases it will now use the right wording. Over time, patterns specific to your office build up (the listing descriptions you use, how you describe an area, your style for directing to a viewing).

For the first two weeks, we recommend spending 10-15 minutes a day on this queue. During this period, the correction rate drops quickly, because the AI learns the frequently made mistakes. Review is the "live" part of training: the system gets smarter as it is used.

7. Open the channels (continuous step)

Once training is ready, you put the assistant in front of customers. intusell gathers all channels in a single inbox: Instagram DM, Instagram comments, WhatsApp, Facebook Messenger, Telegram, Web Chat, and email.

Channel-opening methods:

• Meta channels (Instagram, Messenger): connect with one-click OAuth.
• WhatsApp: connects in about 1 minute by scanning a QR code from your phone — no Meta Business approval required. The official Cloud API option is also available.

In real estate, the first two channels to open are usually WhatsApp and Instagram DM, because the questions coming in under portal listings and posts flow there. Instagram comment automation, on the other hand, is rolled out gradually subject to Meta approval; the DM gate (instagram_dm_enabled) is separate from the comments gate (instagram_comments_enabled). So you can first automate DMs and keep the comments under listing posts manual, then bring those on as the comments gate is rolled out.

There are working modes for handing off to a human: ai_only (the AI answers everything), human_only (everything goes to the agent), hybrid (the AI normally answers and escalates to the agent when needed). Most offices start with hybrid; the AI handles listing-question and viewing traffic, and conversations requiring negotiation and a contract go to the agent. For the details of Instagram and WhatsApp automation: Instagram and WhatsApp lead automation for real estate.

How long does training take?

A working setup takes half a day:

1. Uploading the portfolio and area information to the knowledge base: 1-2 hours (shorter if your Excel portfolio is ready)
2. Property viewing appointment types and working hours: 20-30 minutes
3. Persona, tone, and first response rules: 30 minutes
4. Connecting channels: 1-5 minutes per channel

But there is no moment of "training is done." Over the first two weeks, as you approve and correct responses in the label review queue, the assistant sharpens to fit your office; keeping the portfolio current and running A/B tests are also ongoing improvements. Setup is fast; mastery is continuous.

What it isn't

Putting intusell in the right category matters, because the wrong expectation leads to the wrong setup.

• It is not a property portal or a listings site. It does not publish listings; it learns your portfolio and runs the communication and the appointment with the customer.
• It is not a title-deed/appraisal/mortgage advisor. It does not guess on matters like the official title deed, appraisal value, or final mortgage approval; it directs these questions to the agent and the relevant institution.
• It is not an intermediary that closes the negotiation. The final price, deposit, and contract are always the work of the agent and the office; intusell is not a party to payment or the contract.
• It is not a bot that invents information. It does not state square meters, a price, or a status it is unsure of; it clearly says it doesn't know and hands off to the agent.

In short: not a bot that gives canned answers, but a real estate sales assistant that represents your office to the extent you train it — while stopping where it doesn't know and handing off to the agent.

Frequently asked questions

How long does AI training take for a real estate agency?

A working setup is completed within half a day: uploading portfolio and area information to the knowledge base, property viewing appointment types and working hours, persona and tone settings, and a few response rules. The real refinement builds up over the first weeks as you approve and correct responses in the label review queue. Training is not one-time; it is continuous.

What happens if the AI states the wrong square meters or price?

The AI uses only the portfolio information you upload to the knowledge base; it does not invent square meters, a price, or a title-deed status it is unsure of — it says it does not know and hands off to the agent. To keep information correct, you simply need to keep your listings and portfolio list up to date; this is the most important trust rule.

Do I need technical knowledge to train the AI?

No. All training is done from the panel; no code, API key, or developer is required. You upload the portfolio and area information as Excel/PDF, enter property viewing appointment types and working hours, and write response rules in plain language. You connect channels with one-click OAuth or by scanning a QR code.

Does the AI create the property viewing appointment on its own?

Yes. The AI creates a viewing appointment based on the listing of interest and the customer's availability, accounting for your working hours and full slots; double-booking the same time is prevented by the system. In complex negotiations or special situations, it hands the conversation off to the agent.

Does the AI answer listing questions that come in overnight?

Yes. It instantly answers, 24/7, the "is this flat still available, how many square meters, is it mortgage-eligible" questions that come in overnight under portal listings, Instagram posts, and WhatsApp, and moves a suitable prospect to a viewing appointment. By the time the agent arrives in the morning, leads are warm and sorted.

Is customer and lead data safe under KVKK/GDPR?

Yes. intusell is end-to-end encrypted and KVKK/GDPR-compliant. Personal data you collect in real estate work — such as name, phone, and budget — is stored encrypted; when you upload past conversation recordings, PII masking and explicit consent are applied. All data is isolated on a per-tenant basis.

Next step

You've trained your assistant; next comes using it in daily operations. The next article in the series explains how to run the intusell you trained on a real workday: overnight listing questions, portfolio matching, the viewing calendar, reminders, and lead follow-up. Continue straight from there: how a real estate agency uses intusell.

This was the real estate main article of our sector-based series. We've built the same structure for other sectors too; for example, for your sector neighbor restaurants you can look at restaurant AI training, for tourism tour agency AI training, and for healthcare clinic AI training.

If you'd like to see it live before starting the setup, use Get a demo and we'll open your panel together in a 20-minute session, or write to hello@intusell.com. For package and quota details, you can check the pricing page, and for other guides, take a look at the all articles list.