AI Agents 8 min read 14 November 2026

AI Agents for Real Estate: 7 Jobs Worth Handing Over

AI agents for real estate answer portal leads in minutes, book viewings and chase deadlines. See 7 jobs, costs and rollout steps. Book a free WizeB call

AI Agents for Real Estate: 7 Jobs Worth Handing Over
Part of the Implementing AI Agents guideHow to Implement AI Agents in Your Business

AI agents for real estate are software workers that pick up the repetitive work around a deal: replying to portal enquiries within minutes, asking the qualifying questions, booking viewings into an agent's calendar, answering the phone after hours, drafting listing copy and chasing the documents and dates between offer and completion. They are not a replacement for the agent who walks the buyer through the house. They are the inside sales assistant, coordinator and night-shift receptionist most brokerages cannot afford to hire. For a brokerage, estate agency or property manager with 10 to 100 people, a first agent usually costs $5,000 to $15,000 to build, runs for a few hundred dollars a month and goes live in three to six weeks.

This post is part of our guide to rolling out agents, which starts at wizeb.com/blog/how-to-implement-ai-agents-in-business. If you are still deciding where to begin, read wizeb.com/blog/first-ai-agent-for-business alongside this one.

The Problem: Leads Arrive When Agents Are Busy

Real estate leads arrive at the worst possible times. A buyer fills in a Zillow, Realtor.com, Rightmove or Zoopla form at 9pm on a Sunday. A seller calls during a viewing. A tenant reports a leak at 2am. The agents who could answer are showing houses, negotiating or asleep, and the lead sits in a CRM inbox until Monday.

Speed is what wins those leads. In "The Short Life of Online Sales Leads" (Harvard Business Review, 2011, hbr.org/2011/03/the-short-life-of-online-sales-leads), researchers who audited 2,241 US companies found that firms contacting a lead within an hour were nearly seven times as likely to qualify it as firms that waited longer. Portal leads are often sent to several agencies at once, so the first useful reply usually gets the viewing. Most brokerages know this and still lose leads, because answering fast at 9pm is a staffing problem, not a knowledge problem.

The same squeeze shows up after the offer. Between contract and close, someone has to track inspection dates, appraisal and mortgage deadlines, missing signatures and chasers to solicitors or title companies. That work is rule-based, deadline-driven and spread across email, which is exactly what an agent handles well.

7 Real Estate Jobs an AI Agent Can Take Over

  1. 1Instant portal and website lead response. The agent reads the enquiry from the portal or your site, replies by text or email within a couple of minutes, asks the questions your agents would ask (budget, timing, mortgage in principle, property to sell) and writes the answers into the CRM. Warm leads get booked straight in; cold ones go into nurture.
  2. 2Viewing and showing scheduling. It offers real slots from the listing agent's calendar, books the viewing, sends confirmations and reminders, and handles the reschedules that eat an assistant's afternoon. Our general guide to the timing side is at wizeb.com/blog/automate-lead-follow-up.
  3. 3After-hours and overflow calls. A voice agent answers the office line when nobody can, takes the enquiry, answers listing questions from your own data (price, bedrooms, parking, availability) and books a callback or viewing. Anything about offers or negotiation goes to a person. Running costs are covered in wizeb.com/blog/ai-voice-agent-cost.
  4. 4Listing descriptions and marketing copy. From the listing sheet, room measurements and photos, the agent drafts the portal description, social posts and email blast in your house style. A person approves every listing, and the agent is told to avoid language that describes who a home suits, which keeps you clear of Fair Housing problems in the US.
  5. 5Transaction coordination. The agent tracks the key dates in each deal, reads incoming emails and documents from Dotloop, SkySlope or your solicitor's portal, flags anything missing or late, and drafts the chaser for your coordinator to send. It does not sign, send funds or change contract terms.
  6. 6Tenant and maintenance requests. For property managers, it takes repair requests by text, web form or phone, asks for photos, sorts emergencies from routine jobs, and creates the work order in AppFolio, Buildium or your UK system with the right contractor suggested. Emergencies trigger an immediate call to the on-call person.
  7. 7Past-client and database follow-up. Most brokerages sit on thousands of past enquiries and clients. The agent works through them in small batches, checks for anything that suggests they might move (a valuation request, a lease ending, a long gap since purchase), and drafts personal check-ins for agents to review and send.

Where to start

Export last month's leads from your CRM with the time each one arrived and the time someone first replied. If more than a quarter waited over an hour, start with lead response: the payoff is direct and the agent only needs read access to listings and write access to the CRM. If your agents reply fast but deals stall before closing, start with transaction coordination instead.

What It Connects To

A real estate agent is useful only if it reaches your data. A typical first build touches three systems: lead sources (Zillow, Realtor.com, Rightmove, Zoopla, OnTheMarket and your own website forms); the CRM (Follow Up Boss, Lofty, BoldTrail, Salesforce or HubSpot in the US; Reapit, Alto or Street in the UK); and calendars plus a texting or phone line. Later builds add the listing feed or MLS data you are licensed to use, the transaction platform and, for property managers, the management system.

Give it the narrowest access that does the job. A lead-response agent needs to read listings, read and write CRM contacts, and send messages from an approved number. It does not need access to client funds, contracts or your email archive. Log every message it sends so a manager can review a day's conversations in ten minutes.

The Rules You Have to Build In

  • Consent for texts and calls. In the US, the FCC ruled in February 2024 that AI-generated voices count as "artificial" under the TCPA, so automated calls and texts need the same prior consent as any robocall. Reply to people who enquired; do not cold-text purchased lists.
  • Fair housing. The agent must never steer, screen or describe homes by protected characteristics. Write that into its instructions, test it with tricky questions, and keep a person approving listing copy.
  • Disclosure. Tell people they are talking to an AI assistant. The EU AI Act requires this for chatbots from August 2026, and it is good practice in the US and UK anyway.
  • Data protection. UK and EU agencies need a lawful basis under UK GDPR or GDPR for every message, and a clear opt-out. Keep personal data in systems you control.
  • Hand-off. Offers, negotiation, legal questions and complaints go to a named person, with the conversation summarised, every time.

CRM Add-On or Custom Agent?

Most real estate CRMs and several lead-gen vendors now sell an AI texting assistant or "AI ISA" add-on, usually priced per agent seat or per lead. For a small team using one CRM and one or two portals, that is the right first step: switch it on and watch what it gets wrong. Add-ons fall short when you need more than one system joined up (lead response that also checks the live calendar and the listing feed), rules a settings page cannot express (different questions for lettings, sales and new homes), or work that is not lead response at all, such as transaction coordination or maintenance triage.

A custom agent costs more up front but you own the prompts, workflows and data, and it works across your stack instead of inside one vendor's. Our guide at wizeb.com/blog/build-vs-buy-ai-agents-2026-decision-framework walks through the decision.

What It Costs and How Long It Takes

  • One job, two or three systems (for example, portal lead response written into Follow Up Boss with calendar booking): about $5,000 to $15,000 and three to six weeks.
  • Several jobs or a voice agent on your phone line, with actions such as bookings and work orders: about $15,000 to $40,000 and six to ten weeks.
  • Running costs: model usage, texting or phone minutes, hosting and monitoring usually come to $200 to $1,000 a month for a brokerage handling a few hundred leads a month.
  • Your team's time: a few hours a week during the build to agree qualifying questions and review test conversations, then about an hour a week reviewing escalations.

Those are WizeB's working ranges for growing firms in the US, UK and Europe, not a price list. The full breakdown, including what pushes the number up, is at wizeb.com/blog/ai-agent-cost-guide.

How to Roll It Out Without Losing Leads

  1. 1Pull 200 to 300 real past enquiries for the job you picked, with the replies your best agents sent. That is your test set.
  2. 2Build the agent and score it against them before any client sees it. Agree a pass mark first: right qualifying questions, correct listing facts, zero promises about price or offers.
  3. 3Run it in draft mode for two weeks during office hours: it writes the reply, an agent sends it. Track how many drafts go out unchanged.
  4. 4Switch it on after hours and at weekends first, when the alternative is no reply at all. Widen to daytime overflow once the numbers hold.
  5. 5Review a sample of conversations every week and feed corrections back into its instructions and test set.

Example: A 25-Agent Brokerage Missing Weekend Leads

This is an illustrative example, but the pattern is common. A residential brokerage in Texas with 25 agents gets about 600 portal and website leads a month into Follow Up Boss. An export shows the median first reply on weekday office hours is 20 minutes, but leads arriving after 6pm or at weekends, about 40% of the total, wait until the next working morning.

They start with after-hours lead response only. The agent texts each new lead within two minutes, asks four qualifying questions, offers viewing slots from the listing agent's calendar and writes everything into the CRM. After two weeks in draft mode, it goes live for evenings and weekends, with any question about offers or price handed to the on-call agent. Two months later they add showing reminders and reschedules. The build costs about $12,000 across both phases and runs at roughly $400 a month including texting. The agents now start Monday with booked viewings and qualified notes instead of a pile of cold leads.

How WizeB Builds Real Estate Agents

WizeB builds custom AI agents, workflow automations and voice agents for growing businesses in the US, UK and Europe. For brokerages, estate agencies and property managers, we start with your lead and deal data, pick the one job where an agent saves the most hours or wins the most leads at the lowest risk, connect it to your CRM, portals and calendars with the narrowest access that works, build in the consent and fair-housing rules, and test it against your past enquiries before launch. You own the prompts, workflows, test sets and accounts. Details are at wizeb.com/services/ai-agents, and phone agents are covered at wizeb.com/services/voice-ai.

Plan your first real estate agent

Book a free scoping call at wizeb.com/services/ai-agents. Bring last month's lead export with arrival and first-reply times, and the CRM and portals you use, and we will tell you which job an agent should take first, what it will cost to build and run, and how fast it can go live. Not sure you are ready? Take the free AI-Native Index at wizeb.com/ai-native-index.

Ready to act on this?

We build exactly what this article is about.

Tell us about your situation — we'll come back with a realistic assessment.