All plays › How real estate agents use AI: 21 real plays with results

How real estate agents use AI: 21 real plays with results

21 real plays, each linked to its source · Updated · Picked and checked by Olga Pearson

How are real estate agents using AI?

Agents and brokers in these plays use AI for listing copy, lead follow-up, prospecting and paperwork, and keep a person on anything a client sees. Missy Plunkett writes a listing description in about 20 minutes with ChatGPT. Mike Renick's AI drafts an email and a text for every lead who comes back to his website, and he approves each one. A commercial broker's AI agent found 624 likely owners in 30 seconds, and call 21 won a listing of about $1M. Results are self-reported unless labeled.

At a glance

OwnerBusinessResultToolAI costSource datePlay
Missy PlunkettRealtor, South Carolina and TennesseeAbout 20 minutes per listing description; saves about 4 hours a week (self-reported, 2024)ChatGPTnot statedJune 2024See the play
Lynea CarverReal estate team, agent network and coaching businessFrom about $200,000 a year in salaries to herself, a virtual assistant and about 57 Claude skills; one recorded listing appointment becomes a listing presentation, MLS descriptions and ad scripts (self-reported)Claudenot statedOct 22, 2026See the play
Mike RenickTeam Renick, Sarasota and Longboat Key, FloridaAn email and a text drafted for every returning website lead, sent only after he approves; 135,000 contacts sorted into smart lists (self-reported)Perplexity Computernot statedJun 16, 2026See the play
Sven AndersenAndersen Realty Group, Boston areaA Claude-built tool lists expired, canceled and for-sale-by-owner listings with phone numbers; $4 million in unseen opportunities, not closed sales (as reported)Claudenot statedOct 4, 2026See the play
Rudy MartinezIndustrial commercial real estate broker, Southern California624 likely owners found in 30 seconds; on call 21 he won a listing of about $1M (self-reported)Terrakotta AI agentnot statedOct 21, 2026See the play
Eric BramlettBramlett Partners, Austin, TexasAbout $2,500 a month in software subscriptions replaced with in-house tools built in Claude Code (self-reported)Claudenot statedMay 13, 2026See the play
Director, Anchor PropertyEstate agency, NottinghamActive sales 35 to 57; listing copy from 90 minutes to about 12 (vendor-reported)Claude + n8nabout 85 pounds a monthJul 28, 2026See the play

Each result is as the owner reported it. AI cost is the owner’s own number. Open the play for the source.

Back office8 plays

  • Real estate and propertyClaude

    About $2,500 a month in software subscriptions replaced with in-house tools at a real estate brokerage

    self-reported

    Eric Bramlett, broker-owner of Bramlett Partners in Austin, Texas, writes that over the last year his brokerage replaced roughly $2,500 a month in off-the-shelf software subscriptions with internal tools built using Claude Code.

    When the Austin real estate market turned, Eric Bramlett audited what the brokerage was paying for. In a May 13, 2026 article he wrote for REALTOR News, he says overlapping software was a quiet drain on budget and agent attention. Over the last year, the team replaced roughly $2,500 a month in off-the-shelf subscriptions with internal tools built in Claude Code. He writes that they are not a technology company, and that the question has shifted from what to buy to whether they should build it first. He does not name which subscriptions were replaced. Other cuts in the same article, including Google Ads and farm mail, were not done with AI and are not part of this result.

    By Eric Bramlett, Bramlett Partners, an independent real estate brokerage in Austin, Texas · Source

  • Real estate and propertyGrok Bot

    $55 a month cut from the business cell phone bill by an AI that emailed the account manager

    self-reported

    Lara Kimbrell says Grok Bot saved her $55 on her monthly cell phone bill. She says it handled everything with her business account manager over email, and all she did was upload her latest bill.

    Lara Kimbrell, founder of CV Capital (commercial real estate debt and equity placement), uploaded her latest business cell phone bill to Grok Bot. The bot took it from there and worked the savings out with her business account manager over email. It is a small, repeatable money play: give an AI agent one recurring bill and let it do the back and forth you never get around to.

    By Lara Kimbrell (@Lara_Kimbrell), CV Capital; commercial real estate capital placement · Source

  • Real estate and propertyGrok Bot

    Paying 150 water bills by hand replaced: about 6 hours of work now takes about 1 hour

    self-reported

    He has over 150 water bill accounts to pay for his property management company. That used to be about a 6-hour job for a human. With Grok Bot it takes about 1 hour: the bot signs in, pays each bill, generates an email with the PDF, and hands it to Hermes, which uploads to tenant ledgers and creates the bill.

    Ray (@sharecashBTC) runs property management and spends 10 to 11 hours a day with Hermes agents on company ops. Hermes cannot do physical UI clicks on water-department sites without an API, so paying 150-plus water bills used to need a second person for about 6 hours. He uses Grok Bot for the browser login and payment step, then Hermes files the PDF to each tenant ledger. He says Grok Bot shines where the Buildium API falls short.

    By Ray (@sharecashBTC), Property management · Source

  • Real estate and propertyGrok Bot

    Typing up appraisal data by hand replaced: 4 hours of data entry cut on one report

    self-reported

    Her Grok Bot Appraiser Assistant prepared its first report for the licensed real estate appraiser to finish, and cut his data entry by 4 hours.

    Mars, a homesteader with a past career in finance, built a Grok Bot Appraiser Assistant for a licensed real estate appraiser's office. It worked up its first appraisal report, which the licensed appraiser then finished, and cut his data entry by 4 hours. The bot then reviews the final report to see what it got wrong and what it needs to learn for next time.

    By Mars (@MarsHomestead), Licensed real estate appraiser's office · Source

  • Real estate and propertyGrok Bot

    CPA paperwork and a half empty deal tracker filled in: about 2 hours and 5 or more hours (his estimate) of work

    self-reported

    His CPA sent a 23 bullet email that would have taken him 2 hours to compile, and Grok Bot completed it while he worked. His tracker of closed deals was missing 50% of the data (square footage, acreage and more), work he says would have taken 5 or more hours (his estimate). He gave it access to the addresses and it was done in a day.

    Gunnar is a commercial real estate broker who sells automotive properties like lube shops. His CPA sent a 23 bullet request, and Grok Bot compiled the answers while he worked. His track record of closed deals was missing half its data, like square footage and acreage. He gave the bot the addresses and the tracker was filled in within a day.

    By Gunnar (@GunnarCREbroker), Commercial real estate broker (automotive properties) · Source

  • Real estate and propertyGrok Bot

    Have a Grok Bot audit your property manager’s statements against your contract

    self-reported

    $17,429.36 in property management overcharges found and wired back. The bot flagged a 32% fee within minutes, checked it against the 2019 contract (25%) and emailed the manager with his OK.

    Ryan Debenham owns rental property run by a property manager, and he hadn’t found time to audit the property statements. He put his Grok Bot on the job. Within a couple of minutes it said a 32% management fee “seems a bit high for the market.” He remembered signing for 25% in 2019, so he had the bot review the contract. It found the rate had quietly changed without notice, calculated the overcharge and, with his permission, emailed the property manager. The money was wired back, and the bot now has a standing job to catch slips like this.

    By Ryan Debenham (@ryandebenham), Rental property owner (also CEO of GRIN) · Source

  • Real estate and propertyGrok Bot

    Let a Grok Bot shop your rental insurance, prep your books and pre-screen tenants

    self-reported

    In one week his Grok Bot saved $3,500 a year on property insurance across his rental portfolio, with more coverage. It also categorized every business transaction year to date for his accountant, and for a rental listing it talked with 34 prospective tenants, summarized their pre-screening answers in a Google Sheet and scheduled 8 tours.

    Alex Porrett is a real estate investor with a rental portfolio. In one busy week his Grok Bot did three jobs in the background. It saved him $3,500 a year on property insurance across his rentals, with more coverage than before. It sorted every business transaction for the year so his accountant had clean books. And when he listed a rental, it handled the inquiries: it got 34 prospective tenants to fill out pre-screening, put the answers in a Google Sheet, booked 8 tours and texted his current tenants from a Google Voice number to find times that worked for them.

    By Alex Porrett (@AlexPorrett), Rental property portfolio (real estate investor) · Source

  • Real estate and propertyAny AI assistant (ChatGPT, Claude, etc.)

    Hand AI your CMAs and deadlines, then book the saved hours with people

    as reported

    One agent got back 12 hours a week, put 8 of them on the phone, and went from on pace for 22 closings to 25. Another agent’s past-client referral closings went from 5 to 9.

    Real estate coaches Tim and Julie Harris found that agents who save hours with AI often lose them to busywork. Their fix: hand off CMAs, deadline tracking and first-draft client updates, then block mornings for past-client coffees, sphere events, FSBOs and expireds.

    By Tim & Julie Harris (real estate coaches), with agents they coach, Harris Real Estate Coaching · Source

Leads8 plays

  • Real estate and propertyPerplexity Computer

    AI-drafted follow-up for every returning lead across 135,000 contacts at a real estate team

    self-reported

    Florida agent Mike Renick says his AI assistant spots any lead who comes back to his website, pulls their history from his CRM, and drafts a personal email and text for him to approve. It also helped organize his 135,000 contacts into smart lists. The episode notes put his production at $15M last year and $21M by the end of April this year.

    Mike Renick, a former General Motors global information officer, runs Team Renick in Sarasota and Longboat Key. When anyone looks at a listing on his website, his website lead service creates a record in Follow Up Boss. His AI assistant, Perplexity Computer, picks it up, reviews that person's history, for example a lead who hasn't visited in 235 days, drafts an email and a text, and pings him on Telegram. He reviews both and picks which to send: "nothing goes live you don't look at." The AI also sorted his database into hot, warm and cold groups and helped build smart lists and emails for them, which he approved.

    By Mike Renick, Team Renick, real estate team in Sarasota and Longboat Key, FL · Source

  • Real estate and propertyTerrakotta AI agent

    Skip tracing that took hours done in one click: a ~$1M listing won 21 calls in

    self-reported

    His AI prospecting agent found 624 likely owners in 30 seconds and pulled their phone numbers from Secretary of State records, work he used to spend hours on. 21 cold calls in, he found an owner looking to sell as soon as possible. He won the ~$1M listing, which went live about 10 days before his post, and calls it his first AI sourced deal.

    Rudy Martinez is a commercial real estate broker. He asked an AI prospecting agent to find industrial owners within 10 miles of his Azusa, California listing who would want to trade up into an 8,000 square foot facility. It returned 624 owners in 30 seconds, then found phone numbers for each one from Secretary of State records. He started dialing, and on call 21 he reached an owner of a 3,100 square foot building 2.2 miles away who wanted to sell as soon as possible. He also uses an AI notetaker on his cold calls.

    By Rudy Martinez, Industrial commercial real estate broker (Southern California) · Source

  • Real estate and propertyClaude

    Have Claude find expired and for-sale-by-owner listings to call

    as reported

    A Claude-built tool that finds expired, canceled and for-sale-by-owner listings, with addresses and phone numbers in a spreadsheet, led to $4 million in "unseen opportunities" the team leader says he didn’t have before.

    Sven Andersen leads Andersen Realty Group, a Keller Williams-affiliated team in the Boston area. Finding expired, canceled and for-sale-by-owner listings by hand takes a long time, so the team hired a college student to build a tool in Claude. Connected to other tools, it uncovers the owners’ addresses and phone numbers and puts them all in a spreadsheet the team can call from. Andersen says it freed his time for talking to people and led to $4 million in "unseen opportunities." He also prepaid Claude for the year for his team, and newer agents use AI to rehearse open-house and buyer or seller conversations.

    By Sven Andersen, Andersen Realty Group (Keller Williams-affiliated real estate team, Boston area) · Source

  • Real estate and propertyGrok Bot

    Have a Grok Bot answer every web inquiry and brief you on the owner before you call

    self-reported

    An owner who got his mailer filled out his contact form. His Grok Bot replied right away, looked up what she owns and briefed him before the call. She had a contract to sign 2 hours after she inquired, adding 6 new doors to manage. He says the bot has paid for itself several times over.

    Chris Twiggs is President and Broker at Blair Allen Property Management in South Dakota, Montana and Arizona. His Grok Bot runs the whole path a lead takes. It helped send the mailer with a QR code to the website, filled out the Google Business Profile, and writes blog posts in his voice. When the contact form came into email, the bot thanked the owner right away and asked for a good time to call. It also drafted a text with the same message in case a call didn’t work. Then it looked up the owner, found what properties she holds, and wrote him a summary so he knew her portfolio before he spoke to her.

    By Chris Twiggs (@ChrisTwiggsPM), Blair Allen Property Management (SD, MT, AZ) · Source

  • Real estate and propertyGrok Bot

    Have a Grok Bot build your weekly direct mail list of rental owners from public records

    self-reported

    His Grok Bot replaced the mailing list company they paid, saving $1,000 a year, and the lists are more than double the length. Every Monday it sends 20 owners to mail, each with a branded rent analysis and a letter ready to sign. Early conversion looks like about 3%.

    Chris Twiggs runs Blair Allen Property Management. His Grok Bot pulls rental addresses from county parcel codes for multifamily and by watching Zillow, Trulia and competitor sites for single-family rentals. It sorts owners by unit count (single, 2-4, 5-20, 20-100, 100+) and filters out owner-operators who would never hire them. Every Monday it emails 20 owners to mail, with a branded rental price analysis, a review of how the rental is marketed now (where it’s posted, photo and copy quality, views, days on market) and a personal letter. He signs each letter in blue ink, the envelope is hand-addressed, and owners with more than 20 doors get FedEx. The bot tracks every mailing in a spreadsheet and sends a follow-up letter 3 months later. It also adds birthday and holiday cards for clients, tenants, team families and vendors.

    By Chris Twiggs (@ChrisTwiggsPM), Blair Allen Property Management (SD, MT, AZ) · Source

  • Real estate and propertyChatGPT + Claude

    Put a lead safety net on your realtor inbox so replies go out in under two minutes

    vendor-reported

    Average digital lead response under 2 minutes, 100% follow-up completion on the live safety-net workflow, and an estimated 6 to 8 hours back per week.

    Melissa Avara kept the personal parts of her brokerage and automated the routine ones: lead capture, follow-up, meeting recaps, and onboarding, with ChatGPT, Claude, Calendly, Zapier, and Fireflies.

    By Melissa Avara, Melissa Avara Realty Group · Source

  • Real estate and propertyClaude

    Run three AI agents so your estate agency can progress 60% more sales

    vendor-reported

    Active sales 35 → 57; ~24 hours recovered per week; listing copy 90 minutes → ~12 minutes; applicant re-engagement 11% → 34%.

    Anchor Property, a six-person Nottingham agency, used Claude plus n8n agents for listing copy, sales progression chasing, and applicant re-engagement, about £85/month to run.

    By Director, Anchor Property (quoted in case study), Anchor Property, Nottingham · Source

  • Real estate and propertyLofty AI

    Keep old buyer leads warm with AI follow-up

    vendor-reported

    Follows up with 50 to 100 leads a month

    A 5-agent team leader in Southlake, Texas uses an AI sales assistant to keep following up with leads he couldn't reach by hand. The team grew from 0 to 30 deals.

    By Justin Nimergood, Top Gun Team, Epique Realty · Source

Marketing and content4 plays

  • Real estate and propertyClaude

    Blog posts from 5 to 6 hours down to under 1 at a real estate brokerage

    self-reported

    Kate Hulbert, marketing director at Bozeman Real Estate Group, says a blog post used to take five to six hours start to finish. With an outline from the admin team plus Claude skills for fact-checking and optimizing, "Now it's done in under an hour." Turning a recorded task walkthrough into a written procedure in Notion now takes minutes.

    After losing a long-time staff member, Bozeman Real Estate Group in Montana struggled to keep its marketing consistent, and its blog went quiet. Over three months the team moved every process into Notion and built Claude projects for each department. For the blog, the admin team writes an outline, a fact-checker skill verifies the draft, and a blog optimizer skill checks formatting, SEO and AI search best practices. For procedures, someone records themselves doing a task, Claude turns the transcript into steps with checklists, and another skill publishes it to Notion. New hires now ask Claude how things are done instead of waiting on a manager.

    By Kate Hulbert, Bozeman Real Estate Group, real estate brokerage in Bozeman, MT (marketing director) · Source

  • Real estate and propertyClaude

    Have Claude turn one recorded listing appointment into your whole listing marketing plan

    self-reported

    A real estate agent who was paying about $200,000 a year in salaries now runs her business with herself, a virtual assistant and about 57 AI agents built as Claude skills. One recorded listing appointment becomes a buyer persona, a 13-slide listing presentation, MLS descriptions and ad scripts.

    Lynea Carver runs a local real estate team, a network of 200 agents in five countries and a coaching business. Last year she was paying ops, marketing, content and video staff. This year she went to just her and one virtual assistant in the Philippines for about 6 months, then added a second. She built Claude skills for the work. River reads her DMs, emails and community posts and suggests replies her VA sends. Her listing machine takes a recorded listing appointment and builds pre-listing research, a 13-part buyer persona, a 13-slide listing presentation, MLS descriptions, coming-soon and agent marketing, and ad scripts with targeting. A voice brain dump on her 10-minute drive home becomes 13 pieces of content.

    By Lynea Carver, Real estate team, agent network and coaching business · Source

  • Real estate and propertyGrok Bot

    Have a Grok Bot edit your listing video in 30 minutes

    self-reported

    A three-minute YouTube video of a $2.2M Southwest Florida listing, edited by a Grok Bot in about 30 minutes using about 3% of his plan’s usage.

    Zane Campbell has edited luxury real estate, music and automotive videos for over a decade. Last week he had his Grok Bot (running Opus 5.5) edit a three-minute listing video and says it came out better than 99% of the listing videos he’s seen.

    By Zane Campbell (@zaneccampbell), Zane Campbell Media (Cape Coral, FL) · Source

  • Real estate and propertyChatGPT

    Write listing descriptions in 20 minutes with ChatGPT

    self-reported

    Saves about 4 hours a week

    A Realtor in South Carolina and Tennessee drafts and polishes listing descriptions and client emails with ChatGPT, cutting hours of writing to about 20 minutes per listing.

    By Missy Plunkett, Realtor, South Carolina and Tennessee · Source

Research and notes1 play

  • Real estate and propertyClaude

    An apartment building underwritten and an offer letter sent in 2 hours with Claude

    self-reported

    Real estate investor Ian McDonald says he used Claude to underwrite an apartment building and "get an LOI in the sellers hands in 2 hours vs the old way where I'd still be crunching numbers and setting up calls." It also caught that the seller paid $0 in interest the most recent year, a sign they likely own it free and clear.

    Ian McDonald buys small apartment buildings. Underwriting used to mean long hours and pro formas he wasn't confident in, especially on expense assumptions and local rules. He used Claude to test scenarios, push back on overly optimistic assumptions, read images to sharpen cost and revenue estimates, answer questions about local laws, and work backward from his target to an offer and letter of intent.

    By Ian McDonald, Apartment building investor · Source

More questions owners ask

How do real estate agents use AI?

For listing copy, lead follow-up and prospecting, with the agent approving what goes out. Missy Plunkett drafts listing descriptions with ChatGPT in about 20 minutes each and says it saves about 4 hours a week (self-reported, 2024). Mike Renick's AI assistant drafts an email and a text whenever an old lead comes back to his website, and he approves each one before it sends (self-reported). Rudy Martinez, a commercial broker, had an AI agent find 624 likely owners in 30 seconds; on call 21 he won a listing of about $1M (self-reported).

Can AI write my listing descriptions?

Yes, as a draft you check line by line. Missy Plunkett's prompt tells ChatGPT to use only the facts she gives and not to say who a home is perfect for, to stay inside fair-housing rules (self-reported, 2024). Lynea Carver turns one recorded listing appointment into a buyer persona, a 13-slide listing presentation, MLS descriptions and ad scripts with Claude skills (self-reported); her play says to get the seller's OK before recording. Anchor Property, a six-person agency in Nottingham, says listing copy went from 90 minutes to about 12 (vendor-reported).

Will buyers and sellers trust an agent who uses AI?

We don't have an owner play with numbers on that yet. What the agents here do is keep a person in every client conversation. Mike Renick says "nothing goes live you don't look at." Lynea Carver's virtual assistant reviews and sends the replies her Claude skill drafts (self-reported). Real estate coaches Tim and Julie Harris say one agent they coach moved 8 saved hours a week to the phone and went from on pace for 22 closings to 25 (as reported).

Which AI is best for a real estate agent?

Pick by the job; no agent here compared tools side by side. Missy Plunkett writes listings in ChatGPT. Lynea Carver, Sven Andersen and Eric Bramlett build with Claude. Mike Renick uses Perplexity Computer with his CRM, and real estate video editor Zane Campbell had a Grok Bot edit a three-minute listing video in about 30 minutes (self-reported). Start with the one that does your most painful weekly job, and keep client data out until you know where it is stored.