A fractional AI Support Department for real estate, 24/7 across buyers, tenants, and viewings.

Pull the last three months of buyer and tenant comms from a typical mid-sized developer or boutique brokerage with three to eight properties or a managed-tenant portfolio. The shape is consistent and brutal. Forty percent of the inbound queue is pre-purchase buyer questions: PSF for the unit, GFA on the floorplan, freehold or leasehold status, available payment schedules, financing options, handover dates, what is included in the unit price, what the maintenance fee covers, whether the developer is offering deposit financing. Twenty-five percent is viewing scheduling and rescheduling: book a viewing for Saturday afternoon, can we reschedule from Sunday morning, can we add a unit to the viewing tour, can the viewing happen in Cantonese.

Twenty percent is post-purchase and pre-handover questions from buyers waiting on off-plan completion: construction milestone updates, payment schedule reminders, when will the snagging walkthrough happen, who is the property manager going to be. Ten percent is tenant requests on the managed portfolio: maintenance requests (AC, plumbing, electrical, appliance), lease renewal questions, rent payment confirmations, neighbor noise complaints, building access issues. The remaining five percent is the genuinely novel questions that need a human broker or a human property manager.

The inbound peak is 9pm to 1am local time for buyer inquiries (cross-border investors writing from Hong Kong, Singapore, Tokyo, LA, Sydney) and 6am to 8am for tenant requests (the maintenance issue they noticed when they woke up). The brokerage office is closed for both peaks. The language mix runs through eight to fourteen languages in a typical month: Mandarin and Cantonese from the Greater China investor cohort, Tagalog from the OFW investor base writing from the Middle East, Bahasa from Jakarta and KL, Japanese from Tokyo investors, Korean from Seoul, Arabic from the GCC cohort, English from the US, UK, and Australia. The front desk speaks two of those languages on a good shift. The rest get Google Translate replies that read polite but uncertain, which trains the buyer to defect to a different brand. For the integrated picture across all four real estate functions, see AI for Real Estate. For the cross-industry support context, see AI Support Department and 24/7 AI Customer Support.

Real estate is the most multilingual function in the buyer journey and the worst-equipped to handle it after hours. A Manila developer fielding inquiries from Hong Kong, Singapore, Tokyo, Los Angeles, and Sydney lives the language problem every day. A Bangkok brokerage takes inquiries on the same off-plan launch in English, Mandarin, Thai, and Russian inside a single week. A Dubai developer fields questions in Arabic, English, Hindi, Mandarin, and Russian inside the first forty-eight hours of campaign go-live. A KL brokerage takes Bahasa, Mandarin, English, Tagalog, and Cantonese on the same listing in the same Tuesday. In every case the buyer base spans eight-plus languages and the front desk covers two or three of those well. The gap is structural and headcount cannot close it.

The agents we deploy speak the languages your buyers speak natively. The inbound language is detected on the first word. The reply lands in that language in the brand tone you have spent years setting, with the right formality register for the culture. A Hong Kong buyer writing in formal Cantonese gets formal Cantonese back. A casual WhatsApp message in Tagalog from an OFW investor in Dubai gets a friendly Tagalog reply with the right developer track record context. A Tokyo investor asking about the PSF for a unit in BGC gets a Japanese reply with the actual PSF, the actual GFA, the actual handover date, and the actual payment schedule pulled from the developer inventory feed. Your brokers and your property managers stay focused on the conversations in front of them in the languages they actually speak. The queue across every other language closes itself.

The agent is multi-property aware from day one. The buyer in Tokyo asking about a unit in the Singapore property gets the Singapore property answer with the right PSF, the right freehold or leasehold status, and the right handover timeline. The buyer in Seoul asking about an off-plan tower in Manila gets the Manila property answer with the right developer track record and the right payment schedule. The agent knows which property the inquiry maps to, pulls the right inventory data, the right developer agreement context, the right local context, and answers in the buyer language. Viewing requests get applied to the right broker calendar with the right property code. Tenant maintenance requests get routed to the right property manager with the full tenant context.

The managed-tenant portion of a real estate portfolio runs on a completely different cadence from the buyer funnel and gets even less coverage. Tenants notice maintenance issues when they wake up, between 6am and 8am: the AC is not cooling, the kitchen tap is dripping, the washing machine drum is grinding, the building lobby keypad is rejecting the access code, the elevator is stuck on floor twelve. They send a WhatsApp message or a portal ticket in their own language. The property manager opens the office at 9am and starts triaging. By the time the work order goes out to the vendor, three hours have passed. The vendor arrives at 1pm or 2pm if the tenant is lucky. The tenant has been waiting since 6am with an AC that does not work.

A multilingual AI Support layer fixes the cadence structurally. The agent receives the tenant message at 6:15am, reads it in the tenant language, triages the issue against the property management knowledge base, opens the work order in the PMS (Buildium, AppFolio, or the in-house system) with the right priority code, routes the work order to the right vendor or in-house maintenance staff, confirms a window to the tenant in the tenant language, and tracks resolution. The property manager opens the office at 9am and finds half the morning maintenance queue already triaged, work-ordered, and scheduled. The tenant who messaged at 6:15am gets a vendor at 10am instead of 2pm.

The compounding effect on tenant satisfaction is large. Tenant retention runs three to five percent higher on portfolios where the maintenance response time drops from hours to minutes. Lease renewal rates climb because the friction that drives tenants out (slow maintenance, language barrier on building issues, lost requests) gets solved structurally. Property managers spend their day on the work the building actually needs (capex planning, vendor management, capital improvements) instead of triaging the maintenance queue all morning. The cross-industry version of this multilingual coverage shape is in AI Support for Hospitality, where the same pattern applies to multi-property hotel groups with overnight guest comms.

11:47pm, Manila developer property: a Hong Kong buyer writes in through WhatsApp in formal Cantonese asking about a two-bedroom unit in tower three of the off-plan development. The agent reads the Cantonese, pulls the unit data from the developer inventory feed, confirms the PSF, the GFA, the freehold status, the payment schedule (twenty percent on signing, ten percent at twenty-five percent construction, ten percent at fifty percent construction, balance on handover), the handover timeline (Q3 2027), and the mortgage broker partnership offering deposit financing. The reply lands in formal Cantonese in the brand voice in thirty-eight seconds. The buyer asks two follow-up questions about the developer track record and the comparable nearby off-plan inventory. Both answered in Cantonese. The buyer books a viewing for the next month when she will be in Manila on business. The viewing lands on the matched broker calendar with the Cantonese conversation thread attached.

3am, Bangkok brokerage off-plan launch: a Korean investor in Seoul wants to register interest in an off-plan tower in the central area and asks in formal Korean. The agent reads the Korean, pulls the launch microsite data, confirms the master plan, the tower profile, the unit mix, the payment schedule, the developer track record, and the comparable inventory. The reply lands in formal Korean in twenty-eight seconds. The Korean investor registers interest, the lead lands in the CRM with the full thread, the buyer language preference is set to Korean for all future comms, and the matched broker gets the lead with the qualification thread translated to English. The next morning the broker has the lead ready with the Korean translation attached when they walk in.

6:15am, Singapore property management portfolio: a tenant in a managed condo writes a WhatsApp message in Mandarin saying the AC in the master bedroom is not cooling. The agent reads the Mandarin, opens a work order in the PMS with the AC priority code, routes the work order to the on-call HVAC vendor for the building, confirms a 10am to 12pm window to the tenant in Mandarin, sends a reminder at 9:30am in Mandarin, and tracks the vendor arrival timestamp. The HVAC vendor arrives at 10:15am, fixes the AC by 11:30am, the tenant confirms resolution in Mandarin, and the work order closes. The property manager opens the office at 9am and finds the AC issue already triaged, work-ordered, vendor-dispatched, and confirmed for resolution.

11am, every property and managed portfolio: the GMs and property managers open the night queue summary. Eighty-two buyer inquiries came in between 9pm and 9am across the developer properties, the brokerage listings, and the managed tenant portfolio. Sixty-seven were closed end to end (PSF and GFA questions answered, viewings scheduled, tenant maintenance work-ordered, off-plan milestone questions answered). Eleven were queued for human review with full property and tenant context. Four were flagged as high-value escalations from cross-border investors with multi-unit interest. The GMs and property managers spend their morning on the fifteen that needed them, not on triaging the eighty-two that did not. For the integrated view across all four real estate functions, see AI for Real Estate. For the cross-industry support context, see AI Support Department and 24/7 AI Customer Support.