Technology
One independent permissioned chain,working with off-chain services.
Blockchain does not replace every business system. It provides verifiable shared records for the identity, licenses, work records, acceptance status and settlement basis that all parties rely on.
See how one order flowsOverall structure
From entry point to independent chain,each layer does one job.
Enterprise customers connect within their existing workflows. Off-chain services handle storage, processing, access and payment. The chain records the identity, license, work and settlement states that all parties rely on.
- Content providersEnterprise appsAI service platforms
- Identity, permissions and API access
- Content and evidence servicesAI task servicesLicensing and delivery services
- Controlled files and indexesModels and compute nodesAccess and payment connections
- Independent permissioned chain and smart contractsRights ledgerWork ledgerSettlement ledger
- Review, governance and evidence export
01 / 06Business entryEnterprise customers can call registration, search, license checks, tasks and delivery within their existing workflows.
02 / 06AccessImage libraries, enterprise asset systems, AI service platforms and review systems connect through APIs and SDKs.
03 / 06Business servicesExecution can be run by the project itself or provided by approved third-party services.
04 / 06Off-chain resourcesOriginal images, contracts, creation records, full reports and sensitive materials go into controlled storage.
05 / 06Independent permissioned chainSmart contracts manage identity and roles, content and evidence references, license conditions, task and acceptance records, and settlement status.
06 / 06GovernanceImportant evidence and business records can be exported as data packages and verification information according to permissions.
Independent permissioned chain
What can be stated precisely, the system enforces.Where judgment is needed, authorized services submit receipts.
ORDIVA plans to build an independent permissioned chain with its own network identity, node admission, contracts and governance rules. The project does not treat issuing an asset on another public chain as a substitute for building this network.
Managed by smart contracts
- Identity and roles
- Content and evidence references
- Rights statement status
- Task and acceptance records
- License conditions
- Settlement status
- Dispute and correction processes
A mature base, R&D on the business
The base layer favors mature blockchain technology. R&D focuses on the rights data model, task records, license enforcement and settlement coordination, rather than building new consensus in the first phase.
Independent nodes, step by step
As partner organizations join, the network introduces independently operated validator and audit nodes step by step. The project running several nodes internally and multiple organizations each controlling their own nodes and keys are governance forms for different stages.
Models do not write the terms
Original contracts, machine-readable conditions and the referenced policy versions stay linked. A model cannot generate new terms on the fly in place of the parties' consent, and cannot change payment or rights status on its own.
Off-chain services
Originals go into controlled storage,the chain keeps only digests and states.
Original images, contracts, creation records, full reports and sensitive materials go into controlled storage. Business databases and index systems support search, access control and daily operations, while the chain keeps the necessary digests and key states.
Kept off-chain
- Original images
- Contracts
- Creation records
- Full reports
- Sensitive materials
Kept on-chain
- Necessary digests
- Key states
Each use licensed separately
Private institutional materials have separate use permissions for search, cross-institution matching, public display and model training. Files submitted for duplicate checks are not automatically added to training data.
Licenses connect to actual access
License status connects to file downloads, API calls or data delivery through the access gateway. License changes affect later controlled access as agreed. Copies that have already left the system fall under the relevant contracts and dispute handling.
Payment through connected channels
Connected payment channels handle funds. The settlement basis on the chain and the channel's payment receipts are reconciled separately. The system does not design bookkeeping as a promise to hold customer funds unconditionally.
Records can be exported
Important evidence and business records can be exported as data packages and verification information according to permissions, reducing dependence on viewing everything in a single front end.
AI infrastructure
Managing content,and delivering services.
Services are organized around clear inputs, processing purposes, quality standards and delivered results. Enterprise customers buy an agreed result and do not need to assemble every model and compute node themselves.
Inside the network
- Identify content types
- Generate search tags
- Find duplicate or related files
- Organize source materials
- Extract key contract fields
- Flag missing or conflicting materials
For customers
- Image classification and dedupe
- Asset search and curation
- Material tagging
- Dataset preparation after licensing
No substitute for rights judgments
AI output serves search, task routing and review, and does not directly replace rights judgments. When no candidate match is found, the system does not claim that originality across the whole web has been proven.
Unknown stays unknown
For content with insufficient evidence, unsuitable models or disputes, the system keeps an unknown status and an entry point for manual handling.
Work receipts
Work receipts record the information needed to review a task, including input versions, model and processing configuration, compute node, output digest and task links.
Service track record
A service track record shows which kinds of tasks a provider delivers reliably, which results were accepted and how disputes were resolved. It serves later purchasing and scheduling first.
Scaling with demand
The network scales with real demand. New GPUs, models or nodes should correspond to specific processing tasks and service capacity, not stand alone as a proxy for project value.
Trust and governance principles
These principles are product mechanisms,not metrics on a page.
The system does not make content registration an automatic grant of copyright, does not make providing compute a way to acquire rights in the input content, and does not make joining a task a claim to all later revenue.
Rights, work and payment managed separately
Each party's rights and payments link to specific statements, licenses, contracts and services actually performed.
Evidence, judgment and status kept apart
Original evidence shows what was submitted. A review opinion shows within what scope a party made a judgment. The current status shows which operations the system allows under which version of the rules.
Data controlled by purpose and role
Compute nodes receive only the materials needed to complete the task. Validator nodes, enterprise customers and reviewers see information according to their roles, and joining the network does not mean access to the whole database.
Clear duties for execution, acceptance and governance
AI service providers are responsible for results, reviewers for their acceptance opinions, and the governance mechanism for members, roles and rule versions. Important changes should be authorized and leave a verifiable record.
Delivery reconciled with actual payment
Access eligibility, delivery completion, meeting settlement conditions and actual receipt of funds are different states. The system shows how they relate and which steps are still open.
Records can be exported and handed over
Important licenses, work receipts and delivery proofs can be exported according to permissions, supporting verification by partners and necessary data migration.