What happens when a public agency designs a matchmaking app seriously?
Identity verification. Proof of single status. An onboarding interview. A profile. A 60-minute daily search window. One application per day. A match. A first meeting. An optional certified volunteer. A “getting-to-know-you” stage of up to 90 days. Then, only if both people agree, “serious dating.” That is when names and phone numbers are finally released.
This is less like a swipe app and more like a surprisingly polished workflow engine for human relationships.
Registration and system use are free. There is an AI recommendation feature. Human help is available when needed.
If you enter expecting “a free dating app,” you may eventually think: “Wait. Did local government build a relationship-state machine?”
This article anonymizes the service completely. No region name, municipality name, service name, address, or personally identifying information appears here.
1. The interview felt more like onboarding than an interrogation
One anonymous user expected questions about values, dating history, and ideal partners.
Instead, the interview was mostly a walkthrough of how the system works.
The staff checked documents and explained search limits, applications, first meetings, relationship stages, and prohibited behavior. At least in this individual case, there was no long personal interview resembling a private matchmaking consultancy.
That fits the product design.
This is not a traditional matchmaking agency where a dedicated counselor regularly chooses candidates and feeds back every interaction. Users mainly operate the service themselves on a phone or computer. Humans step in for identity checks, onboarding, questions, and optional first-meeting support.
Normal mode: self-service. Exception mode: human escalation.
It is basically matchmaking SaaS.
2. Sixty minutes a day, one application a day, three concurrent cases: dating has rate limits
Search and profile browsing are limited to one continuous block of up to 60 minutes per day. Logging out does not pause the clock. The allowance resets at midnight.
You may apply to only one person per day. Once you send that application, search and browsing end for the day even if you still had time left.
There is also a combined cap of three active matches or “getting-to-know-you” relationships. When that cap is reached, new search and applications stop, and your own profile is hidden from other members.
A commercial app may encourage: “Browse 100 people, like 20, message 10.”
This system says: “Please deal with one human at a time.”
User: “Okay.” System: “You already applied today. The store is closed.”
It is dating with an API rate limiter.
The design intent is obvious: introduce friction against mass applications and force more deliberate choices.
3. The AI watches behavior, not magic — but the weights are a black box
Recommendations refresh on a fixed day each week and can include up to three people. If there is not enough behavioral data, there may be no recommendation.
Official materials explicitly say that profile-view history and application history are used as behavioral signals.
So browsing is not necessarily meaningless.
Who you look at and who you apply to can become recommendation data.
The FAQ also says that people who have applied before may be more likely to appear as a recommended candidate to members of the opposite sex.
But that does not justify the leap: “One application gives me a permanent algorithm boost.”
The service does not publicly disclose:
- how far back history is weighted,
- whether old actions decay,
- the relative weight of views versus applications,
- whether weekly activity matters,
- whether recent login affects ordering,
- how long any application-related exposure effect lasts.
So applying to people you do not actually like just to “feed the algorithm” is a poor strategy. It may not work, and it may contaminate the preference data the system is trying to learn.
Do not date for the robot.
4. No confirmed visible footprints. But the AI still sees the footprints
On many dating services, profile “footprints” are powerful.
You view someone. They see that you viewed them. They visit back. A low-cost signal turns into an application.
In this anonymized service, the public FAQ and user guides do not confirm a visible “who viewed me” feature or a ranking based on last login.
Yet the system explicitly records viewing history for AI recommendations.
So the situation is almost comical:
Other user: “I can’t see who looked at me.” Algorithm: “I can.”
Visible footprints are not confirmed. Internal behavioral tracking is.
That distinction matters.
5. If you want to be mostly passive, the profile becomes the product
For people who hate grinding through dating apps, a mostly passive strategy is possible.
But total abandonment does not fit the recommendation design as well as occasional genuine activity.
A low-effort version looks like this:
- Build the profile properly once.
- Keep photos public if comfortable.
- Do not overuse hard deal-breaker filters.
- Once in a while, browse only people you genuinely find relevant.
- Apply only when you actually want to meet.
- Otherwise, wait.
Think 95% waiting, 5% real activity.
In one staff interview, the service was described as having roughly a thousand members, around 70% of them men. That was not a published statistic and is included only as an anecdotal reference. If you belong to the larger side of an imbalanced pool, being completely passive is usually less favorable simply because more people are competing for fewer counterparts.
Still, there is no need to behave like a day trader.
The main advantage of a free service is that holding the option open costs almost nothing.
6. “Non-negotiable” filters are powerful enough to make you disappear
Users may designate up to three preferences as non-negotiable.
This does not merely hide incompatible people from your own search.
It can also prevent your profile from appearing to people who do not satisfy those conditions.
And if a non-mandatory profile field becomes part of a hard filter, users who left that field blank may also stop seeing you.
That creates a funny failure mode:
“Great, irrelevant applications disappeared!”
followed by:
“Wait. I disappeared too.”
This is especially important for passive users. If you plan to wait for inbound interest, over-filtering can kill your own visibility.
There is also a heart-based compatibility display tied to the other person’s stated preferences. It tells you how well your profile fits what they say they care about.
The hearts are not romance.
They are basically a conditional JOIN result.
7. The volunteer supporter is the star of the system
The most memorable feature is a certified volunteer who can support the first meeting.
After a match, both users answer whether they want a supporter present.
If either person requests one, the supported route is activated.
That creates the following scene:
Person A: “Can’t we just meet?” Person B: “I want a supporter.” System: “Supporter summoned.”
Each participant pays a small fixed amount for that supported meeting. Official rules describe it as covering the supporter’s travel and food-related expenses, so it is not literally “free lunch money.”
The supporter coordinates a concrete time and place from the submitted availability. Before the “getting-to-know-you” stage, direct chat between the two members is not opened in the supported route; communication runs mainly through the supporter.
On the day, the supporter checks membership credentials and introduces the two people.
Then the supporter leaves early.
This is where the NPC energy becomes impossible to ignore.
“Hero A, meet Hero B.” “My work here is done.” Poof.
The two people then continue talking on their own for about an hour and report completion afterward.
If the relationship proceeds to the next stage, the supporter can remain available for advice.
The tutorial NPC has a persistent support channel.
8. No LINE, no personal contact exchange, until the serious stage
After the first meeting, both people decide whether to enter the “getting-to-know-you” stage.
Despite the word “dating,” this is not yet treated as an exclusive romantic relationship.
It can last up to 90 days. The pair may meet repeatedly, but communication stays inside the platform.
Until “serious dating” is established, exchanging a real name, home address, phone number, email address, LINE account, or similar direct contact information is prohibited.
On a normal app, the flow is often:
Match ↓ A few messages ↓ “Want to switch to LINE?”
Here the system says:
“Not yet.”
The trade-off is clear. Users lose some freedom, while the operator retains an auditable communication channel longer and can intervene if something goes wrong.
9. Serious dating is the unlock screen for names and phone numbers
Within the 90-day stage, if both people agree that they want an ordinary romantic relationship, they can enter “serious dating.”
Only one serious partner is allowed.
When that status begins, other searches and parallel processes stop, and the profile becomes unavailable to new candidates.
Then the operator releases each person’s name and phone number to the other.
In videogame language:
Achievement unlocked: Serious Dating. Item acquired: Real Name. Item acquired: Phone Number. Feature unlocked: Off-platform contact.
The guidance also explains that the service checks relationship status again at later points rather than disappearing immediately.
The public agency is surprisingly committed to follow-up.
10. Is the structure restrictive, or is it a safety system?
Both readings are reasonable.
The advantages are straightforward:
- free entry and use,
- identity and single-status verification,
- less mass-spam behavior,
- delayed exchange of direct contact information,
- optional third-party help with the first meeting,
- human support when needed,
- a limit on parallel relationships.
But some users will find it restrictive:
- short daily search time,
- one application per day,
- slow path to direct messaging,
- a supporter can be triggered by one side’s preference,
- the supported route reduces freedom over logistics,
- the whole process feels strongly rule-governed.
That does not mean “people who use supporters are bad matches.”
Someone who wants third-party reassurance and someone who wants two adults to decide everything directly may simply have different process preferences.
That is a compatibility issue, not a character ranking.
11. This is not really a free dating app. It is a relationship workflow with customer success
The cleanest description may be:
A local government combined a dating interface, identity verification, a recommender system, case-state management, and optional human support.
A normal app mostly optimizes getting two people into contact.
This service manages the stages after contact as well:
Search. Application. Match. First meeting. Getting-to-know-you stage. Serious dating. Contact-information release. Follow-up. Marriage or closure.
Dating has become a state machine.
Ironically, that rigid structure can be convenient for people who dislike spending time on dating apps.
Build the profile once. Use a public photo. Do not over-filter. Occasionally browse people you truly like. Apply only when someone is genuinely interesting. Then wait.
There is no need to worship the algorithm.
Visible footprints are not confirmed, but browsing data feeds the recommender. A past application may increase recommendation exposure, but the duration and weight are unknown. So this is neither “one click gives lifetime priority” nor “log in every day or die.”
For a low-effort user, the sensible model is probably not total inactivity but “mostly wait, occasionally generate honest signals.”
And one final note about the volunteer supporter.
The role is serious and legitimate.
But from the pure mechanics, it still looks exactly like an NPC who is summoned by one player, receives a small expense payment from both sides, introduces the party, and leaves the scene.
Once you see it, you cannot unsee it.
Evidence level and anonymization
This article is based on official service guides, FAQ materials, terms of use, the onboarding guidance available as of 2026-09-30, and one anonymous user’s interview experience.
To preserve complete geographic anonymity, all municipality names, region names, service names, addresses, URLs, and geographic identifiers have been removed.
The approximate member count and roughly 70/30 gender split came from an anonymous staff conversation and are not treated as official published statistics. The service does not publicly disclose recommendation weights, history windows, decay functions, last-login ranking, or a visible footprint feature, so this article does not claim any of those mechanisms exist.
