AI-backed self-representation is cutting buyer costs in Florida - and testing the old commission model
Florida buyers are quietly closing on homes without agents, using AI to find listings, price offers, draft contracts, and submit within hours. Homa, a startup built by former Zillow product leader Arman Javaherian, says at least 10 homes have already closed end-to-end on its platform, with more in escrow.
The timing isn't random. After the National Association of Realtors settlement pushed buyers to sign contracts before touring, the true cost of buyer representation is now in plain view. Some buyers are opting out - and keeping the 2.5% to 3% that used to fund their agent.
Read the NAR settlement context
Real savings buyers are reporting
DJ, a 32-year-old pharmacist in Tampa Bay, bought a $420,000 home through Homa and kept $10,500 that would have gone to a buyer's agent. "They pretty much walk you through the whole thing step by step," he said. "You can end up saving at least 2.5% on your final purchase price."
Vicki Lynn, who moved to Vero Beach from California, purchased a $313,000 home and saved about $8,000 - enough to cover a significant slice of closing costs. "The contract system was similar to TurboTax - filling in the blanks. Very straightforward."
How Homa works
Homa combines search, instant tour scheduling, AI pricing analysis, and automated contract creation. Buyers can tour fast and submit offers the same day.
There's a free self-service flow, plus a $1,995 "Homa Pro" tier that assigns a licensed transaction broker and guarantees the buyer receives the rebated commission. Without that, some listing agents try to keep both sides of the commission when an unrepresented buyer shows up.
Why this matters to real estate and construction pros
- Commission compression: If buyers self-represent, listing agents will field more direct inquiries and face scrutiny on co-broke payouts.
- Speed as a feature: Offers generated in hours raise the bar for response times, disclosures, and coordination with lenders and title.
- Listing strategy: Expect more direct-to-listing communication and requests for rebates or credits tied to buyer self-representation.
- Operational risk: Automated contracts still require tight compliance, accurate timelines, and clean addenda. Sloppy process will surface fast.
- Local intel premium: Off-market context (upcoming projects, HOA issues, assessment risk) is a clear edge for pros who can articulate it.
- Builder/direct sales angle: New-home teams can capture AI-ready buyers by simplifying tours, spec inventory data, and incentives.
The counterpoint from brokers
Not everyone's convinced AI can replace the human element. Miami broker Ivan Chorney argues that true representation hinges on context that isn't in public records: motivation, pressure, ego, timing. Local intel - from future development to unpublicized construction - can swing value and isn't captured by models scanning listings.
His view: the best outcomes come from professionals who use AI to enhance judgment, not from AI trying to replace it.
Autonomy with oversight
Javaherian is not pitching a free-for-all. "All AI has hallucinations," he said. That's why Homa Pro assigns a human transaction broker to check offers and negotiations.
The goal: 95% to 98% autonomy, not 100%. The biggest purchase of someone's life still needs a safety net.
Where buyers are already using AI (with or without agents)
- Redfin: Plain-language search to surface neighborhoods that fit lifestyle constraints.
- Zillow: Automated price estimates to gauge whether a list price feels fair before calling anyone.
- Opendoor: Algorithm-driven pricing and timelines in speed-sensitive markets.
- Flyhomes: Q&A tools that answer basic questions buyers may hesitate to ask a human.
What to do now
- Make speed and clarity your edge: publish offer instructions, timelines, and required docs on every listing.
- Codify local intel: maintain a living brief for each listing (upcoming projects, noise, insurance history, HOA status) and share on request.
- Offer unbundled services: paid consults, negotiation-only, inspection-walkthrough packages. Meet buyers where they are.
- Tighten compliance: standardize addenda, contingency timelines, and escrow coordination to handle same-day offers.
- Train teams on AI tools: search prompts, pricing checks, and contract QA to keep pace with AI-accelerated buyers.
What's next
Homa is currently live in Florida and plans to expand into Texas and California next year. The company is backed by Restive Ventures and Cambrian Ventures and is raising its next round.
Whether AI stays niche or becomes a mainstream alternative is still an open question. But as commissions face new scrutiny and more buyers try self-representation, the pressure on traditional agent models is building. As Javaherian puts it: "Most people don't even realize they're allowed to buy a house without a buyer's agent - until now."
Want to upskill your team on practical AI workflows? Explore role-based programs at Complete AI Training.
Your membership also unlocks:
