ES Contact
Peer-to-peer platforms

Software for peer-to-peer platforms: exchanges that close themselves

Two strangers agree on an exchange and the platform sees it through without manual mediation.

Let's model your operation
The starting point

There's no operator to assign the task to here, but the task still exists.

In a peer-to-peer marketplace there's no operator to assign the task to: the actors are the users themselves and the algorithms that match them. But the work exists all the same, and something has to orchestrate it before a problem shows up.

Capsules

Your information

The things that exist in your business, each one in a single place.

Listing

Item, condition and terms.

User

Verification, reputation and history.

Exchange

Parties, agreement and guarantees.

Meeting point

Location, coverage and hours.

Review

Outcome and evidence.

States

Your stages

Each stage stops being a label and starts carrying its own rules.

Exchange agreed

Time limit
72 hours to arrange the meetup.
Entry / exit triggers
Enters when both parties accept; exits with a place and time set.
Data validation
Both users are verified.
Dynamic permissions
Only the two parties involved.

Meetup pending

Time limit
The date agreed between the parties.
Entry / exit triggers
Enters with the meeting point chosen; exits with the handoff confirmed.
Data validation
The meeting point is within the covered area.
Dynamic permissions
The parties and, if it escalates, support.

In dispute

Time limit
Seven days to resolve.
Entry / exit triggers
Enters when one party reports an issue.
Data validation
Both parties have attached evidence.
Dynamic permissions
Support, with the payment on hold.

Closed

Time limit
No limit: it's the final state.
Entry / exit triggers
Enters when both parties confirm.
Data validation
Review submitted and payment released.
Dynamic permissions
Automatic.
Glup · Aimee

Who does the work

The same work, split between people, agents and machines depending on the moment.

  • HumanUsers
  • AI agentMatching algorithm
  • ServiceIdentity verification
  • HumanSupport
  • AI agentAimee
How it works, with your case

An exchange between two users who've never met

  1. 01

    Entry

    The exchange enters «Exchange agreed» when both parties accept.

    Exchange
  2. 02

    Trigger

    The stage triggers Glup with the 72-hour clock running.

    Glup
  3. 03

    Context

    Aimee checks both users' reputation, the distance, and whether either has a history of disputes.

    Aimee
  4. 04

    Delegation

    High reputation: direct meetup. Risk profile: verified meeting point and payment on hold.

    Algorithm or support
  5. 05

    Close

    Once the meetup is confirmed, the exchange closes and reviews are requested.

    Glup

What changes

How it works today

Support steps in when something has already gone wrong and there are two angry users.

How it works with Astrolle

The level of protection is decided before the meetup, based on the context of each exchange.

AI isn't here to replace your team

It's here to decide who handles what, and when, so people keep the part only they can do: judgment, relationships and exceptions.

Let's model your operation

Change industry

Retail and logistics Agribusiness Appointment-based services Construction and real estate Education Peer-to-peer platforms

Don't see yours? It works the same way. Write to us and we'll look at your specific case. Let's talk about your case

The next step starts here

Contact us