A
AI-managed group conversations
Persistent, supervised chat threads with built-in context, summarisation, and clarification — running natively on the channels participants already use.
LPGP.network turns the slow, fragile, fully-manual relay of anonymous professional dialogue into a structured, AI-assisted protocol — designed to preserve confidentiality while removing the friction that has always defined this kind of work.
A single, supervised pipeline that adapts to the sensitivity of each conversation rather than forcing every interaction through the same template.
Participants — whether principals or intermediaries — onboard through an invitation. We verify them once, give their organisation a starting balance of access credits, and assign them an opaque identity used for all subsequent dialogue.
Rather than describing a request in free-form prose, participants compose an enquiry through a guided template. The system uses that structure to find the small number of counterparties for whom the conversation is genuinely relevant.
Anonymous group conversations run on familiar messaging surfaces — most often WhatsApp. An AI assistant moderates, summarises, and answers what it already knows; participants only see what concerns them, and never each other's identifiers.
Before any party is formally introduced, the system helps both sides articulate and align the terms they actually care about — structurally, not as a negotiation transcript. Disagreements surface early and cheaply, instead of breaking the deal late.
When the conversation is ready — and only then — identities are revealed to the consenting parties. A live introduction is scheduled. From that point onwards, LPGP steps back; the relationship belongs to the participants.
Persistent, supervised chat threads with built-in context, summarisation, and clarification — running natively on the channels participants already use.
Every message is rewritten and re-sent by the system in a form that preserves meaning without leaking the originator. Names, firms, and signals are revealed only by explicit consent.
The system does not blast enquiries across the network. It engages only the small set of participants for whom the conversation is materially relevant.
Before any introduction, the platform helps both sides express, compare, and reconcile the terms they actually care about — quickly and without back-channel chatter.
A small, accountable human team reviews edge cases and steps in whenever a conversation requires judgement that should not be left to a model alone.
Every dialogue is retained, scoped, and queryable by the participants who own it. Nothing happens off the record; nothing is shared beyond it.
If you are coordinating the kind of dialogue this page describes, we'd like to hear from you. We onboard a small number of organisations each quarter.