Real Estate Agent AI Stack 2026: Listing Lens + EVA + Vmake (30-Min Setup)

Three real-estate-specific AI tools launched in seven days. Listing Lens AI by ProspectsPLUS (April 29) gives every U.S. single-family home a “Sellability Score” so you can predict which homeowners will list. EVA by HomeLight (April 27) is the industry’s first AI escrow officer — it does most of the 120 closing tasks on its own. Vmake’s AI People Remover (April 27) erases passersby and your own car from listing photos in one click.

That’s a real change to the agent’s daily stack. The combined time saved per active listing, on a regular week, is meaningful — easily 4-6 hours if you’re carrying 8-12 listings. But you have to set them up in the right order, and only one of the three is worth paying for past the trial. Here’s the 30-minute walkthrough.

What changed this week (and why it actually matters)

Most “real estate AI” coverage in the past 18 months has been about content tools — listing copy, captions, social posts. Useful, but not deep. This week’s three tools land on the actual workflow: who to prospect, what the photo looks like, and how the file closes.

Tool	Launch date	What it replaces	Time saved	Free trial
Listing Lens AI (ProspectsPLUS)	Apr 29, 2026	Manually building geographic farm lists	2-3 hrs/week	Built into ProspectsPLUS! plan
HomeLight EVA	Apr 27, 2026	Coordinator handoff + 120-task closing checklist	8-12 hrs per closing	Per-transaction pricing, varies
Vmake AI People Remover	Apr 27, 2026	Photo re-shoot + manual editing	30-60 min per listing	Free tier with watermark

Two of the three (Listing Lens, EVA) are software tied to broader platforms; one (Vmake) is a standalone tool. None of them require you to learn anything new about AI. If you’ve used ChatGPT to write listing copy, you already know enough.

The interesting question isn’t what they do — it’s which order to set them up in. Which is the part the press releases skip.

The order matters: predictive → photo → close

Most agents who try multiple tools at once end up using none of them. The trick is sequencing — set up the upstream tool first, then the next stage, then the close. Each stage feeds the next.

Stage 1 — Predictive farming (Listing Lens AI): Find sellers before they list. Stage 2 — Listing photos (Vmake): Clean up the photos so they actually attract those sellers. Stage 3 — Closing automation (EVA): Handle the file when offers come in so you can keep prospecting.

Here’s how each stage actually works.

Stage 1: Listing Lens AI — predictive farming

ProspectsPLUS has been the legacy direct-mail farming platform for residential real estate for years. Listing Lens AI is the first time their stack got predictive. Every U.S. single-family home now has a Sellability Score — a probability that the homeowner will list within a defined window. The score updates as inputs change (life-event signals, equity position, days-on-market in the local micro-market, etc.).

What you do (5 minutes):

Log into your ProspectsPLUS! account. If you don’t have one, start a 14-day trial.
From the dashboard, find the “Listing Lens AI” tab (it’s the third item, between Geographic Farm and Lists).
Select your geographic farm boundary (your existing one, or draw a new one).
Set your target window — most agents start at 90-180 days.
Click “Generate Farm List.” The tool returns the top 50-200 homes ranked by Sellability Score.

What you get is a CSV-equivalent list of addresses, owner names, predicted Sellability Score, and the underlying signal (e.g. “high equity + 8+ years owned + recent permit pull”). That’s the input for the next step — your weekly outreach.

The “is it worth it” question is the harder part. We wrote a separate 4-question audit on whether to pay for the Sellability Score — the short version is: if your average commission is over $9K and your geographic farm is over 200 homes, the math usually works.

What it can’t do: Predict timing precisely. It tells you who’s more likely to list — not when. Plan a 12-week outreach cycle, not a one-shot.

Stage 2: Vmake AI People Remover — listing photo cleanup

Listing photos are the reason most properties get a second look. Vmake’s new “Passerby Mode” (launched April 27) selectively removes background individuals while keeping the main subject intact. There’s also “People Mode” that removes everyone from the frame — useful for busy walkthroughs, less useful for the standard living-room shot.

The two cases this fixes that previously needed a re-shoot or paid edit:

A passerby walking through frame in a curb-appeal photo
Your own car (or seller’s car) in the driveway

Both are common. Both used to mean either re-shooting or sending the file to a $25-50 photo editor.

What you do (10 minutes for 10 photos):

Go to vmake.ai/ai-people-remover (no install needed; works in browser).
Upload up to 50 photos at once. Vmake’s batch mode handles them in parallel.
Tag the people you want removed. The “Passerby Mode” auto-detects most.
Review the AI-regrown background. Vmake fills with generative content matching color, grain, and light.
Download full-resolution photos.

The free tier watermarks your output. The paid tier is around $20-40/month depending on volume — light by listing-photo standards.

What it can’t do: Remove people standing very close to a wall (the regrown wall texture can be visibly off). Walk-throughs with three or more people in a single frame can produce visible artifacts. Test on one photo before batching.

Stage 3: HomeLight EVA — AI escrow agent

EVA is the most ambitious of the three. HomeLight launched it as the industry’s first agentic escrow officer — meaning it doesn’t just template documents, it actually executes work across 80+ tools. Opening orders, ordering HOA docs and title, interfacing with lenders, wiring funds. The full 120-task escrow checklist gets handed off, with EVA flagging the items that need your judgment.

The financial backing is real. HomeLight raised $40 million in debt financing from BlackRock specifically to scale EVA nationwide. That’s not the kind of capital you raise unless the underlying product is working.

What you do (15 minutes for one practice closing):

Sign up at HomeLight Closing Services.
Add EVA to one practice file — pick a transaction that’s at the inspection stage.
EVA introduces itself to your buyer’s agent and the lender via standard email handoff.
Watch what EVA does in the next 48 hours. The dashboard shows every task: ordered, completed, flagged for human review.
Review the “flagged” queue daily. Anything EVA isn’t sure about goes to you.

The key insight from beta testers: EVA handles the busy work, but you still own the relationships. The handshake conversations — buyer concerns, lender hiccups, seller stress — are still you. EVA just stops you from doing the document chase that used to fill the gap between handshakes.

What it can’t do: Replace a transaction coordinator on a complex commercial deal. Replace your judgment on contingency contestation, repair credits, or anything that requires reading the room. EVA is best on standard residential transactions; the further you get from that, the more flagged-for-human items you’ll see.

What this means for you

If you’re a solo agent doing 6-12 transactions a year:

Start with Stage 2 (Vmake). It’s the lowest-cost, fastest-payback tool. Before you decide on Listing Lens or EVA, you’ll save 30-60 min on every listing photo set just from removing passersby and parked cars. After 30 days, evaluate Listing Lens against your current farm list. Skip EVA for now — at 6-12 transactions, the per-transaction cost calculation rarely beats your existing TC arrangement.

If you’re a 1-3 person team doing 20-50 transactions a year:

Run all three. Vmake for photo cleanup is a no-brainer. Listing Lens makes sense if you’re geographic-farming over 200 homes weekly. EVA hits hardest when you’re carrying 4+ active files at once — the time-back compounds because you can prospect more without falling behind on coordination.

If you’re at a brokerage thinking about training the team:

The bigger question isn’t which tool to roll out — it’s whether your agents understand which stage they’re optimizing. Newer agents tend to over-optimize photo cleanup (it feels productive). Mid-career agents skip predictive and miss the upstream gain. The sequencing — predictive first, photo second, close third — is the part to teach.

If you’ve never used AI tools as a real estate agent:

Start at Stage 2. Photo cleanup is the cleanest “show me the result” workflow. Get one listing’s photo set done with Vmake; see the time saved; then move upstream. Don’t try all three in week one.

What none of them do (yet)

A few honest gaps in the 2026 real estate AI stack:

Buyer-side prospecting at the same depth. Listing Lens predicts sellers; the buyer-side equivalent doesn’t exist with the same data quality yet. Buyer-finding is still LinkedIn, referrals, open-house data.
Multi-MLS data normalization. Listing Lens works on national single-family-home data. If you’re cross-MLS-shopping (e.g. dual licensure across state lines), you still need separate tools.
Closing on commercial or unique residential. EVA is built for the standard residential transaction shape. Anything off-pattern (estate sales, probate, REO, FSBO with no agent on the other side) needs a real coordinator.
Compliance with state-specific disclosure rules. Especially Texas, California, and New York — each has wrinkles that need your judgment, not EVA’s. Don’t assume EVA’s standard checklist matches your state.

The 30-minute setup checklist

If you’re starting from scratch this week, here’s the exact 30-minute sequence:

Minutes 0-10: Sign up for ProspectsPLUS! free trial → enable Listing Lens AI → generate first farm list (50 homes minimum)
Minutes 10-15: Note your top 10 Sellability Score addresses for next week’s outreach
Minutes 15-25: Sign up at vmake.ai/ai-people-remover → batch-clean your last 10 listing photos to learn the tool
Minutes 25-30: Start a HomeLight EVA practice file on one of your active inspection-stage transactions — let EVA handle the next 48 hours of coordination, review the flagged queue tomorrow morning

Total cost for the first 30 days: ProspectsPLUS! trial (free), Vmake free tier ($0), HomeLight EVA per-transaction (varies, often $50-150 per practice file — call the rep for current pricing). Less than $200 for the month, even on the high end. Compare that to your hourly rate × 4-6 hours saved per active file — the payback math is straightforward.

For the deeper “should I keep paying for Listing Lens past the trial” decision, see our 4-question audit. For the broader real-estate-AI training, our free 8-lesson AI for Real Estate Agents course covers the prerequisites these tools assume.

The bottom line

Three real launches in seven days, all profession-locked, all with measurable time-back. Vmake first (10 minutes, mostly free). Then Listing Lens for the agents whose farm justifies it. EVA last, and only if your transaction volume makes the per-file math work.

Don’t try all three this week. Get one stage working before you add the next. The agents who try the whole stack at once usually quit two weeks in. The agents who add one stage every two weeks compound through Q3.

Sources:

Real Estate Agent AI Stack 2026: Listing Lens + EVA + Vmake (30-Min Setup)

Table of Contents