Welcome to AutomateRE, your weekly playbook for using AI in real estate.
Big week in AI and Real Estate. HomeLight just launched EVA, an AI escrow agent that automates most of the 120 tasks required to close a residential transaction and hit 100% accuracy across four major workflows by May. Backed by $40 million in new financing from BlackRock, this is not a beta product. It is live, it is scaling nationally, and it is the clearest signal yet that the back office of real estate is being automated from the bottom up. At the same time, a buyer spent five months trying to purchase a home using only AI and no agent. The experiment did not go the way most people assume.
This is issue #18. Let's get into it.
Quick Win: A full monthly market report, neighborhood by neighborhood, with your commentary baked in ready to send before the first of the month
Here's the prompt:
You are a real estate market analyst writing a monthly report for a specific
agent's client list. Your job is to turn raw market data into a readable,
opinionated report that sounds like it came from someone who actually
works in this market every day.
MY PROFILE:
- Name: [FIRST NAME]
- Market: [CITY, STATE]
- Speciality: [BUYERS / SELLERS / INVESTORS / NEIGHBORHOODS]
- Tone: [e.g. "direct and data-driven" / "warm and local" / "straight talk"]
MY MARKET DATA THIS MONTH:
- Median sale price: [$X and vs. last month / last year]
- Median days on market: [N days and vs. last month / last year]
- Active listings: [N and vs. last month / last year]
- Months of supply: [N]
- List-to-sale ratio: [X%]
- Price reductions this month: [N or %]
- One neighborhood to call out: [name + what happened there specifically]
MY COMMENTARY (paste rough notes even messy is fine):
[Anything you observed: a bidding war, a price that shocked you, a
neighborhood that went quiet, a buyer type you keep seeing, a
seller who waited too long]
BUILD THREE SECTIONS:
1. THE NUMBER THAT MATTERS THIS MONTH
- Pick the single most important data point from above
- One paragraph: what it means for buyers, what it means for sellers
- No hedging. Take a position.
2. NEIGHBORHOOD SPOTLIGHT
- Use the neighborhood I flagged above
- What happened there this month, specifically
- What it signals for the next 60 days
- One practical implication for anyone thinking about buying or selling there
3. WHAT I'M TELLING MY CLIENTS RIGHT NOW
- Three bullet points maximum
- Written in first person your voice, your read on the market
- The kind of thing you'd say over coffee, not in a press release
- End with one forward looking sentence: what to watch next month
FORMAT:
- Subject line for the email version
- Total length: 350-450 words
- No charts, no tables this is a reading experience, not a dashboard
- Sign off with agent name and one line: market, phone, reply invitation
TONE: [AGENT TONE FROM ABOVE]. You have an opinion. You are not
a data aggregator. You are the person who was in three houses
last week and knows what the numbers actually mean on the ground.Result: Raw market data and two minutes of rough notes become a 400 word report your clients will forward to their neighbors ready to send before the first of every month.
Industry Intel: AI is automating the transaction itself not just the lead
Two stories this week, and they both point at the same shift. AI is moving past the front end of the deal. It is now inside the transaction, handling the work that used to require a coordinator, a title rep, and three hours of back-and-forth email.
HomeLight launched an AI escrow agent and it just hit 100% accuracy
On June 4, Inman reported that HomeLight launched EVA, an AI-powered escrow agent designed to automate most of the 120 discrete tasks required to close a residential real estate transaction. Opening orders, pulling HOA documents, interfacing with lenders, wiring funds EVA handles them. By the first quarter of 2026, EVA hit what HomeLight defines as 100% accuracy across four major workflows, reached between February and early May. HomeLight announced $40 million in new debt financing from funds managed by BlackRock to scale the platform nationally. CEO Drew Uher told Inman: the company has now automated about 25% of the escrow surface area and has line of sight to automating most of it.
Your move: The back office is being automated faster than most agents are tracking. That is not a threat to your business it is a compression of the transaction timeline that makes your value clearer, not smaller. An agent who closes faster, with fewer errors, and with a market report that lands on the first of every month is the one clients refer. The prompt above is the front end of that reputation. EVA is the back end. Together they look like a seamless experience.
A buyer spent five months trying to purchase a home with only AI. Here's what actually happened.
On June 3, Inman published the account of a buyer who spent five months walking the homebuying path using only AI tools and no agent. The experiment surfaced the gap that matters: the neighbor down the street who planned to sell agreed to a showing only because of a human connection made through the grapevine not a portal, not an AI search, not a Zestimate. The deal that almost happened required someone the buyer trusted. That person was not an algorithm.
Your move: AI handles the research, the data, the drafts, and now the escrow. What it does not handle is the neighbor who calls you because she trusts you specifically. The monthly market report above is how you stay that person. Not by being smarter than the AI. By being present when the AI is not enough.
Automate This: Monthly Market Report on Autopilot
Every week, we help you automate one part of your job. This week: a workflow that writes your monthly market report, neighborhood by neighborhood, with your commentary baked in and sends it to your entire list before the first of the month.
The problem: It is the last week of the month. You know you should send something to your list. You have the numbers somewhere a spreadsheet, a brokerage email, a MLS report you downloaded and never opened. Writing the report means finding the data, forming an opinion, writing 400 words that do not sound like a press release, formatting it for email, and hitting send before everyone else sends theirs. So you do it late, or you send the brokerage template, or you skip it entirely. Your name stays out of your clients' inboxes for another month. Someone else's does not.
The solution: We built you an automation workflow for n8n:
On the 25th of every month, a form triggers automatically on your phone median price, days on market, active listings, months of supply, one neighborhood to spotlight, two minutes of rough notes from the field. Five minutes to fill.
The workflow pulls your saved agent profile from a settings tab: name, market, tone, speciality.
It sends everything through the Quick Win prompt above in one call.
AI returns a fully formatted market report: subject line, three sections, your sign-off 400 words that sound like you wrote them after being in three houses last week.
The report lands in a Google Doc for your review. You read it, change one line if needed, approve it.
The workflow sends it to your full client list via Gmail or Mailchimp on the first of the month at 7 AM before anyone opens their laptop.
Everything logs to a Report Sheet: month, open rate, reply count, any listing or buyer inquiry that came from it.
After six months, you have a record of which market narratives generated the most responses. You stop writing about what happened and start writing about what your clients actually react to.
Time to set up: 35–45 minutes
Time saved: 90–120 minutes per month, every month
Want it? Reply to this email with "MARKETREPORT" and we'll send you the workflow file + setup guide.
Next week: an AI that prepares your full buyer consultation package market positioning, neighborhood comparison, offer strategy, and a financing reality check before you sit down with a new buyer client.
That's it for issue #18. HomeLight automated escrow. A buyer tried to go it alone with AI for five months and learned what you already know. The market report above is how you stay the person clients call when the algorithm is not enough. Reply if you want the workflow. And we'll see you next Tuesday.
- NextAutomation Team
PS: Know another agent, investor, developer or PM who'd find this useful ? https://newsletter.nextautomation.us/subscribe - it helps us keep this free.