Inside the engine · for the ones who need more than a promise

No black box. Six layers, each one inspectable.

Real Estate Hero is the 24/7 lead management system for South African estate agencies. The overview page shows you what it does. This page shows you how, layer by layer: what catches the enquiry, what talks to the buyer, what decides, what your agents see, what you see, and where the humans sit in all of it. If you're the kind of principal who asks "but how, exactly?", this page is for you.

Layer 01

Capture. Nothing gets in without being caught.

Every enquiry route your agency has, wired into one timeline. Not "most of them". All of them, and we verify each one during the build by enquiring ourselves and watching it land.

Portal enquiries

Property24 and Private Property enquiries arrive as emails to an address the engine watches. Each one is parsed into a contact with the listing attached, seconds after the portal sends it, instead of sitting in someone's inbox until Monday.

WhatsApp on a business number

Listing buttons point at a registered WhatsApp Business number owned by the agency, not anyone's personal SIM. Every message lands on the record. Agents keep their own numbers for their own relationships.

Facebook and Instagram

Lead ads, Messenger and DM enquiries flow in through Meta's own integration. If you run property ads, the person who taps them gets the same 60-second treatment as a portal lead.

Missed calls become messages

A missed call to the office line triggers an immediate text back: "Sorry we missed you, how can we help with the property?" The call that used to ring out at 18:05 now starts a conversation.

Your website, every form

Whatever your site runs on, its forms and chat land in the same place. We do not rebuild your website, we catch what it produces.

One buyer, one record

The buyer who enquires on Property24 on Saturday and WhatsApps on Tuesday is recognised as one person, one timeline. No duplicate calls from two agents. That alone has embarrassed agencies out of deals.

Verified, not assumedDuring the build we send test enquiries through every single route and watch them arrive. You get a checklist with a tick per channel, dated. If a route can't be wired for some platform reason, you know before go-live, not in month three.
Layer 02

Conversation. The voice that answers at 22:47.

An AI assistant, trained on your agency, that opens every conversation in under a minute. Not a menu bot, not "press 1 for sales". A conversation a buyer actually continues, and you approve its voice before it ever meets one.

Trained on your agency

Your areas, your stock, your office hours, your tone, the phrases your best agent uses. Built during onboarding, then tuned against real conversations in the first weeks. You sign off the voice before launch.

Answers about the actual listing

The enquiry arrives attached to a listing, so the first reply is specific: the price, the area, availability for viewings. Specificity is why buyers keep typing instead of ghosting.

What it will not do

It does not negotiate price, does not give legal or bond advice, does not invent answers about a property. When it does not know, it says so and brings in a human. Guardrails are part of the build, not a promise.

Escalation, immediately, for the ones that matter

A seller asking for a valuation, a complaint, anything sensitive: flagged and routed to a human straight away, with an alert. A mandate lead never gets the chatbot treatment.

The honest answer on "is it a bot?"If a buyer asks directly, the assistant says it is the agency's digital assistant and offers a human. In practice buyers care about speed and useful answers at 22:47, not about who typed them. The conversation on the overview page is the standard we tune to; if a reply would embarrass your brand, that is a build fault and we fix it in the tuning weeks.
Layer 03

Qualification and booking. Homework before handover.

Four facts gathered in conversation, not in a form: budget, area, timeline, finance status. Then the viewing goes straight into a diary. Your agent's first contact with the buyer is a confirmed appointment, not a cold call.

Asked like a person, not a form

"So I only show you homes worth your Sunday, what range are you comfortable with?" gets an answer where a Budget dropdown gets abandoned. The four facts arrive as conversation, and land as structured data on the record.

Real diary slots

The engine offers actual availability from the assigned agent's calendar, books the slot, sends the confirmation and the reminder. Reminders are why the Sunday viewing actually happens.

Unqualified is a result too

The dreamer with no bond and a five-year horizon is not discarded, they are marked, nurtured monthly, and kept away from your agents' mornings. Six months later, when the bond is approved, they are already yours.

Handover with context

The agent receives the appointment with the full conversation, the four facts, and the listing. Nobody asks the buyer to repeat themselves, which buyers read as competence.

Layer 04

Routing and follow-up. Nothing dies quietly.

Every enquiry is assigned to a named agent, timestamped, and chased by the system until something happens. The lead that used to die after one unanswered call now either converts or gets flagged to you with a name on it.

Routing rules you choose

Area specialist first, listing agent first, round robin, or any mix per office. Agreed during the build, changeable any month. The rule is written down, so lead allocation stops being office politics.

Follow-up until an outcome

No reply after the viewing? The sequence follows up on day 1, day 3, day 7, in your voice, until there is an answer or a clear dead end. The second contact is where deals come from, and it is the number most agencies never measure.

The cold ones surface

An enquiry an agent has not touched inside the agreed window gets flagged: first to the agent, then to you. Quietly ignoring a lead stops being invisible, which is most of the cure.

The pipeline stays when agents leave

Every conversation, every contact, every outcome belongs to the agency's system. An agent resigning takes their relationships, not your record of every buyer who ever enquired.

Layer 05

Intelligence. The layer that reads everything so you don't.

Hundreds of conversations a month produce more information than any principal can read. This layer compresses it: summaries, drafted replies, suggested tasks, and a score on every lead.

Smart summaries

Every conversation compressed to three lines, and every week compressed to one paragraph: what moved, what stalled, what pattern is forming. You read the story, not the transcript.

Smart messages

When a human reply is needed, a draft is waiting, written in your agency's voice, specific to the thread. Agents send, tweak or scrap it. The blank-page delay disappears.

Suggested tasks

Stalls become to-dos with names on them: "Karin hasn't replied in three days, nudge her." "Johan has four leads with no second contact." The system remembers so nobody has to.

Predictive prioritisation

Every lead scored on the four facts plus behaviour: reply speed, viewing attendance, message depth. Mornings start at the top of the scored list, not the top of the inbox. The score's inputs are visible, never a mystery number.

What feeds the scoreBudget vs stock, finance status, stated timeline, and observed behaviour. A pre-approved buyer who replies in minutes and booked a viewing outranks a bigger budget that has gone quiet for a week. You can open any score and see why it is what it is.
Layer 06

Visibility. What you see that nobody else does.

Agents see their own work and nothing else. You see everything: every office, every agent by name, every enquiry's fate. Two surfaces carry it: the owner view for your mornings, the Owner Report for everyone you answer to.

The owner view

The live workspace: scored queue, per-agent performance, source performance, the week's summary. Two minutes with coffee. It is our workspace configured for your agency, with an owner-level view your agents don't get, and we say so plainly.

The Owner Report

One page, per office, per month, as a PDF branded to your agency. Enquiries by source, median first reply, qualification rate, viewings booked, the per-agent league, reviews banked, and second contacts vs died-after-one. Built to be forwarded to partners who will never log into anything.

Portal spend becomes a decision

When you can see that Property24 produced 96 enquiries and your site produced 42, and what each converted, next year's portal negotiation happens with numbers in hand.

Reviews, credited to the agent

Every closed deal triggers a review request credited to the agent who earned it. The league table makes review-collection a competition, and your Google profile compounds while competitors stall.

The layer under all of it

The humans. This is managed, not handed over.

The monthly fee does not buy software access. It buys people at JCE Media running the machine: tuning conversations, watching the queues, assembling the report, training your agents. If the engine misses, that is our phone that rings.

Weeks 1 to 2 · Diagnostic

We read your agency

Live enquiry-path audit, response baselines, review standing per town. You get the findings report and the four numbers. Nothing installed yet, nothing disrupted.

Weeks 3 to 5 · Build

One office goes live

Channels wired and verified, voice trained and signed off by you, routing rules agreed, agents onboarded one by one, day-one baseline captured. Personalised onboarding means each agent is shown their own workspace with their own leads, not a webinar.

Ongoing · Managed

We run it, you read it

Conversation tuning, queue monitoring, monthly Owner Report, support and training as your team changes. Offices two and onward join on the pilot's evidence, cheaper, because they inherit the build.

A modern South African home at dusk with its lights on

You've now seen more of this engine than most owners ever see of their own lead flow.

That's the point. Every layer above is inspectable, measured and on the record. Bring the hard questions below to the walkthrough; they're the ones we enjoy.

The hard questions

Asked by the sharpest principals we've sat with.

The overview page answers the common questions. These are the harder ones, answered the way we answer them in the room.

Who owns the data if we leave?
You do. Contacts, conversation records and outcomes export in standard formats and go with you. Leaving can't mean losing your pipeline, so the exit is part of the design, not a concession we make reluctantly.
Why not just hire an admin or an inside-sales person instead?
Hire one if you can find one who works Saturday 22:47, never takes leave, answers eight channels at once, writes everything down, and costs from R9,500 a month all-in. The honest comparison: a person does empathy and judgement better; the engine does speed, coverage and record-keeping better. Most agencies need the second list fixed first, and the engine hands your existing people better mornings rather than replacing anyone.
Our franchise head office mandates a CRM. Does this clash?
Usually no, and we check before we build, in writing. The engine handles the first minutes and the follow-up discipline; your mandated CRM stays the system of record if head office requires it, and we feed it. If a franchise rule genuinely forbids third-party lead handling, we would rather find that in the diagnostic than after you've paid for a build, which is another reason the diagnostic comes first.
What happens when the AI gets something wrong?
Three things, in order. The guardrails keep it away from the expensive mistakes (price negotiation, legal advice, invented facts). The escalation rules bring a human in the moment it is unsure or the topic is sensitive. And every conversation is on the record, so when it does say something clumsy, you or we see it, and it gets tuned that week. Compare that to the mistakes currently happening on personal phones, which nobody sees and nobody fixes.
Will my agents actually use it?
The engine does its job whether or not agents love it, because capture, answering and booking don't depend on agent adoption. What needs adoption is the workspace, and that is why onboarding is personal: each agent is set up with their own pipeline and shown their own leads on their own screen. The league table does the rest. By month two the same agents who grumbled are checking their standing, because visibility cuts both ways: the good ones finally have proof.
POPIA, in more detail than a reassurance?
Consent is captured at the point of enquiry, with the buyer initiating contact. Data is held in one access-controlled system: agents see only their own leads, the owner view is yours, exports are logged. Deletion requests are actionable because the data is in one place, which is precisely what the personal-handset status quo cannot do. If your attorney wants specifics before signing, we'll walk them through it; that conversation has never lost us a deal.
What does cancelling actually look like?
It's month to month. You give notice, we hand over your data exports, the WhatsApp business number stays with the agency, and the report templates are yours to keep using manually if you want. No exit fees, no held-hostage anything. We keep clients by being the thing that works, not by the contract.

Convinced enough for twenty minutes?

The walkthrough is the layers on this page, live, on your own listings. We open one together and you watch your current enquiry path the way a buyer experiences it. Then you decide about the diagnostic.

Re-read the full offer

Book your walkthrough

We reply within one working day. Usually much faster. We are, after all, the speed people.

Booked in. Talk soon.

We've got your details and will come back within one working day with times. Check your inbox.