SunderSystems

Answer every lead, book the viewing, post the listing.

An AI front office for real estate teams. It drafts replies to buyers and renters, proposes viewing times, writes listing posts and follows up. You approve from your phone, or let routine tasks run on their own. AI requests go to an open-weight model: the untuned 4B on our hardware, or a larger one through OpenRouter.

  1. Mara, Messenger, 11:42 pmHi, is the 2-bedroom at Sunrise Residences still available? Can I see it this weekend?
  2. Draft reply, ready for approvalHi Mara, yes, Unit 8B is still available: 2 bedrooms, 54 sqm, 3.2M. I can show it Saturday at 10:00 or 14:00. Which suits you?
  3. Owner, 7:05 am, on a phoneApprove and send.
  4. After Mara picks 10:00Booked: Saturday 10:00, Sunrise Residences Unit 8B. Calendar invite sent to Mara and to you.

Sample with an invented listing. Messenger replies are planned. The same flow runs on email in our demo today.

Where agencies lose leads.

  • Leads write at night

    A buyer messages on Facebook at 11 pm. By the time someone replies the next afternoon, they have booked a viewing with another agency.

  • Slow replies lose viewings

    The first agent to answer with a time usually gets the viewing. Waiting hours to check a calendar costs you that race.

  • Listing posts slip

    New listings need photos, copy and posts on each channel. When the team is busy showing homes, the posts wait.

  • Follow-ups never happen

    Leads who asked once and went quiet get forgotten. A short, timely message would bring many of them back.

What the front office does.

Every message it writes waits for your approval unless you tell it a task can send on its own.

  1. Reads the lead

    It picks up new email and looks up the listing the lead is asking about.

  2. Drafts a reply with the facts

    It answers with the price, size and availability from your listing, in your tone.

  3. Proposes viewing times

    It checks your calendar and offers slots that are actually free.

  4. Asks you to approve

    You get the draft on your phone, edit it if you like, and send it with one tap.

  5. Books and follows up

    It adds the viewing to the calendar and sends reminders. Messaging leads who go quiet is in pilot.

Listing posts and reports too.

Social posts for each listing
It writes posts for Facebook and Instagram from your listing details. You approve the copy and the schedule.
A weekly report
Leads received, response time, viewings booked and what each channel brought in.
Channels
Email and calendar work in our demo. Live connections to your own inbox are new; your own calendar is planned. Facebook and Instagram posts are drafted for your approval; publishing straight to your pages is planned. Messenger, Viber and WhatsApp inboxes are planned.

Buyer and client data stays private.

Where client data goes
Your workspace runs on our hardware. AI requests go to an open-weight model: the untuned 4B model on our hardware, or a larger open-weight model through OpenRouter. Requests sent through OpenRouter leave our hardware and go to OpenRouter and the provider it routes to. With the 4B, a request does not leave our hardware.
You decide what leaves
Every connected account and every model choice is yours to change or switch off.
Your own model is planned
Per-client training is in development. When it launches on Growth and Private, a model trained on your conversations will be yours to export.

See it in a real operation.

HousingInteractive uses an AI agent team for communication, scheduling, outreach and content. HousingInteractive is operated by our founder. Its agents currently use cloud models; moving them to a model trained for it, on our hardware, is the next step.

Plans start at $99 a month (₱5,700) plus usage. Sign-up is by invite code for now.

Ten founding places, open to any industry.

Lock your price for 12 months, get done-for-you setup and a direct line to the founder.

See the founding programme