Solutions

Who this network is built for.

Ranked by fit rather than ambition. Autonomous agents come first because the problem is sharpest there and the primitives map onto it most directly. The rest follow in order.

StatusActive development. Each segment names the subsystems it depends on and the ones that are unfinished, before it describes what is possible. What works today.

Priority 1 · Where we start

Agents need authority, not just intelligence.

You are building autonomous software that has to authenticate, hold permissions, reach services privately and pay for what it uses. Today those four answers live in four systems, and the agent owns none of them.

Identity

An AI agent is a first-class identity type, distinct from a personal identity and a node-operator identity, carrying a supervisor relationship to a human identity.

Authority

Rights are delegated with a scope, a spend cap and an expiry, and revocation cascades to everything derived from the delegation.

Reachability

A service is reachable at an address decoupled from the machine hosting it, so an agent can call it without learning where it runs.

Payment

A signed transfer between two identities, settled inside the delegated cap with no human approval step in the path.

Infrastructure for agents Read the architecture

In preparation

Two segments the architecture maps onto as directly.

Written material for these is in preparation. The primitives they rely on, and the status of each, are already published on the architecture and build status pages.

Priority 2 · In preparation

DePIN and decentralized infrastructure

You are building a compute, storage or network protocol, and you have started writing the five tables you did not set out to write: providers, reputation, payments, wallets and permissions. Identity, settlement and capacity reservation are network primitives here rather than schema your team maintains. Capacity discovery covers CPU, memory and storage, not GPU-aware scheduling, and multi-node orchestration is still evolving.

Priority 3 · In preparation

Private infrastructure

Your problem is not encryption. It is everything visible around the encryption: which service talks to which machine, and what the operator hosting you can observe. UNP is a routing architecture, not an anonymity guarantee. Hardware-backed confidential computing is not implemented, so an operator with root on the host is inside the trust boundary, and none of it has been independently audited.

Also being designed for

Priority 5 · In preparation

Autonomous organizations

Organizational chains define ownership, treasury and cryptographically authenticated voting across several governance models. Significant parts of the external interface remain placeholder or mocked, so this is a design conversation rather than a migration path for a live treasury or a live vote.

Priority 6 · In preparation

Sovereign infrastructure

Long-horizon by declaration. There are no certifications, no independent security audit, no SLA and no production references to offer. If you are running a tender this year, UniNet is not a bidder. The ask here is a design conversation, not a pilot contract.

Supply side

The other half of the network is the machines.

Datacenters, VPS providers, GPU owners and node operators host chains and isolated workloads on hardware they already run. Capacity is allocated as shared or dedicated, and each operator identity carries a reliability score recorded at the network layer rather than inside one marketplace database.

Reward economics are not finalized, and there is no live consumer demand on the network yet — capacity discovery is a primitive, not a queue of paying jobs.

Run a node