Reppo Subnets
Subnets in Reppo are lightweight, self-governing environments where knowledge, skill, and creativity can be published, evaluated, and rewarded.
Anyone can start a subnet. It's like setting up your own data collection workspace.
Public Subnets
Public subnets are the open layer of Reppo. They’re where anyone can share work, join discussions, and help surface valuable insights in real time.
Anyone can join. There are no gatekeepers or steep setup costs. Contributions can come from individuals, teams, or even AI systems.
Voting defines value. Community members stake tokens and vote for what’s credible, useful, or original.
Signals become data. Every vote adds to a growing dataset of human preferences — a living record of what people across cultures and fields find meaningful.
The cycle sustains itself. Preference data is monetized by the Reppo Foundation, with proceeds cycling back into buybacks, rewards, and ecosystem support.
Public subnets are where collective human intelligence is mapped and measured in the open.
Private Subnets
Private subnets are for groups that need focus, privacy, or higher signal. They can be set up by enterprises, research teams, DAOs, or individual developers — with almost no friction.
Low barrier to launch. Spinning up a subnet is simple and affordable. You don’t need massive capital or compute.
Custom logic. Each subnet sets its own rules — who can participate, how evaluation works, and how rewards are shared.
Expert-only or hybrid. Subnets can be closed to verified experts or open to collaborative human–AI systems.
Full data control. The subnet owner decides who sees the data and how it’s used — ideal for fields like medicine, law, or finance.
Private subnets make it easy to create high-quality, purpose-driven human feedback on proprietary source data.
Accessible by Design
Reppo’s subnets are built to be simple, modular, and inclusive. Starting one shouldn’t feel like applying for a grant or configuring a data center — it should feel like starting a conversation that pays everyone who adds real value.
Launch cost
Low — no major staking or compute required
Who can create
Anyone: individuals, teams, enterprises
Governance
Defined by each subnet
Languages supported
70+
Output
Clean, structured preference and expertise data
The Bigger Picture
Reppo’s subnet architecture is designed to make AI trraining data composable — not locked inside institutions, paywalls, or black-box models.
Each subnet adds a new piece to the global map of expertise, making the AI training data eventually into a knowledge graph.
Last updated