Use case · Growth · Sales
Matchmaking
AI matchmaking matches one side of your market to the other automatically: buyers to properties, candidates to roles, investors to opportunities. New criteria are matched against all your stock straight away, new listings in one daily run, and every match becomes a deal record for your team to work. Nobody has to scroll a spreadsheet to find it.
- Qualified matches in three months, one UK brokerage
- 400+
- Deals in negotiation from that platform
- 67
- From new buyer lead to matched listings
- < 60s
What is AI matchmaking?
AI matchmaking is an automation that compares structured criteria on one side of a market with structured listings on the other, filters on the hard requirements, and uses AI to judge the close calls from free-text details. Each match creates a deal record linking the two sides, with a match ID, so your team can follow it up.
Why do good matches get missed?
Good matches get missed because matching is manual. A new listing comes in and someone scrolls a spreadsheet of buyers' briefs; a new buyer signs up and someone searches the stock by eye. When the person who remembers the right buyer is busy or away, the match never happens, and nothing records who was matched to what.
Matching depends on memory
The right buyer is found only if the right person happens to remember them.Only one brief gets checked
Buyers with several different briefs get only one of them looked at.Five listings, five emails
When listings arrive together, the same buyer gets a stream of separate messages in a day.
How does AI matching work?
AI matching filters on the rules that matter, such as type, budget and location, then an AI step reads the free-text details on both sides and judges the close calls. In practice, one AI step writes the database search from each brief and a second step checks every result against the brief before anything is sent. New criteria match instantly; new listings match in a weekday batch.
With AI automations
New buyer criteria, a listing becoming available, a scheduled daily run or a manual match
Structured criteria
Each buyer, candidate or investor has one or more criteria sets, each matched separately.Structured listings
Each property, role or opportunity is stored with the same fields.Rules first, then AI
Every hard requirement must fit; an AI step judges the free-text details.Only match who's ready
Only Active contacts and Available listings are matched automatically.The right timing
New criteria match against all stock at once; new listings match in one scheduled run, so buyers get one round of matches, not a stream.Every match becomes a deal
A deal record links both sides, with a match ID and type (AI or Manual).Notify the right person
A teaser email to the buyer or their agent, or a deal in your pipeline to send by hand.Manual match
Your team can match any listing to a chosen contact in one click.
For a UK property deal-sourcing brokerage, new listings are matched against every active buyer in one batch on weekdays at 9:00 AM UK time, and buyers are only matched once the team has verified them and set them to Active.
How do you keep both sides confidential?
You keep both sides confidential with status gates, anonymised teasers and filtered views. Matches can be shared as AI-written teasers that leave out the address and anything else private, and each confidentiality rule gets its own filtered view, so buyers don't see addresses, sellers don't see buyers and partners only see their own clients until your team makes the introduction.
- Status gates. Nobody is matched until your team has verified them.
- Teasers. A short summary of each match that holds back identities and addresses.
- Filtered views. Every rule has a matching view: buyers don't see addresses, sellers don't see buyers, partners see only their own clients.
Which systems deliver it?
Matchmaking is delivered by AI automations. There's no fixed platform: it can run on Airtable with Make.com or n8n, or as a custom build with Supabase or Convex as the database. Your data can stay where it is, or move into a structured database we set up for you, with teasers and a deal pipeline built alongside.
| System | Trigger | What it does | Package note |
|---|---|---|---|
| AI Automations | New criteria, new listing, daily run, manual match | Rules filter plus AI judgement, deal records, teasers and notifications | Airtable with Make.com or n8n, or a custom build on Supabase or Convex |
Where is it used?
Matchmaking suits any business that earns its fee by connecting the right people on two sides of a market. Real estate calls it property matching; recruitment calls it matching candidates to roles; commodity traders call it matching buyers and sellers, or sourcing; and placement consultancies match clients to the right programme or provider.
Same job, their words
Property matching: a buyer's brief for commercial property under a set budget is checked against every available listing the moment it's entered, and new listings are matched each weekday morning. Each match becomes a deal, and the buyer gets a teaser without the address.
Proof
400+ qualified matches in three months
400+ qualified matches in three months · 67 deals in negotiation · < 60s matching
From spreadsheets to one portal with AI matching
Weekday 9:00 AM UK matching batch · only verified buyers matched · handed over September 2026
We also built a demo for a commodity trading business: supplier listings and buyer requests are matched from both sides by commodity, with AI picking the best matches on quantity, budget and timing.
What are the limits?
Matching is only as good as the criteria and listing data behind it: garbage in, garbage out. We agree the required fields with you and make the forms enforce them. Setting a contact to Active is deliberately left to your team, so nobody is matched before they've been verified.
- Data quality decides match quality. Required fields are enforced at entry.
- Status is a deliberate step. Your team activates contacts after checking them.
- AI judges, people introduce. Matches go to your team or out as teasers; introductions stay with you.
Frequently asked questions
Questions
Answer
How does AI decide what's a good match?
AI matching first applies the hard rules you agree, such as type, budget and location, and every one must fit. Then an AI step reads the free-text details on both sides, like a buyer's notes or a listing description, and judges the close calls. A second AI step checks each result against the brief before anything is sent.
Can buyers see the property address?
Not until you decide they should. Matches can go out as AI-written teasers that leave out the address and anything else confidential, and buyer views in a portal show the area, category, price and teaser only. Your team releases full details when it makes the introduction, usually after an NDA is signed.
Why match new listings in a daily batch?
So buyers get one round of matches a day instead of a stream of separate emails. New listings are matched in a single scheduled run, for example on weekday mornings. New buyer criteria are different: they're matched against all your stock straight away, because a new buyer wants to see what's available now.
Can a person override a match?
Yes. Your team can match any listing to a chosen contact in one click, which creates the same deal record marked Manual. AI matches can be reviewed in your pipeline before anything is sent, and setting contacts to Active stays with your team, so you control who is matched at all.
Does it work outside property?
Yes. The same logic works for any two-sided market: candidates and roles in recruitment, buyers and suppliers in commodity trading, clients and programmes in placement consultancies, and investors and opportunities in funding networks. The fields and rules change; the approach doesn't.