Fluence brings the next generation of internet computing, which is permissionless, scalable, and composable.
Anyone can connect to the network, run a node, deploy services and applications without any permission of a third party. Applications can leverage existing services to build faster and provide a better user experience. Node operators can re-host existing services providing better resilience under a higher load.
Workloads deployed to the network are globally discoverable by their immutable identifiers and can be executed by anyone as long as it follows workloads' access control policy. API-version support hell is eliminated by permanent service and function addresses and opt-in providers.
Any external data source, API, or another decentralized protocol can be plugged into applications on Fluence. This allows usage of Fluence as a decentralized computing layer on top of various Web 3 protocols, decentralized storage, blockchains, DID, etc.
At its core, Fluence leverages IPFS as a default data layer for distributing and updating services on network nodes.
Applications will be able to create as many subnetworks in the global Fluence network as they need, so the global capacity is estimated in billions of peers. Peers in subnetworks have better connectivity among each other and prioritize outside connections lower. This way the network scales via shards and their ad-hoc communication.
Various workloads can be announced and discovered over the Fluence network. Applications may offload tasks to external service providers who are known by their public credentials or using decentralized identity standards and verifiable credentials. These providers may serve required API, computation; provide a certain type of hardware (e.g. GPU, Trusted execution environments, etc).
The security model of Fluence allows applications to execute only on nodes they know and trust. All execution is cryptographically signed, so clients can always check if the computation was performed by an eligible peer.
In case when trustless consensus is required, it can be deployed for a specific job. The optionality of consensus enables applications to combine blockchain-level security with a greater scale of trusted computations.
Fluence supports various types of network consensus as Aqua libraries. Because Aqua is perfect for implementing network coordination algorithms, both trusted (e.g. Raft) and trustless (e.g. Tendermint) consensus engines can be developed on Aqua and integrated into applications.