global AI agent market cited in the BP
Blockchain-native AI execution infrastructure
LCEN powers verifiable AI execution.
A Light Compute Execution Network for agent identity, verifiable training, multi-agent strategy factories, confidential validation, and on-chain workload settlement.
- 5
- execution layers
- CU
- workload ledger
- TEE
- validation path
EXECUTION LOOP
LIVE
- 01Model Meshrouting + model BOM
- 02Agent Runtimeidentity + task control
- 03Strategy Factorytraining + simulation
- 04Validation GatewayTEE + zkML + replay
- 05On-chain Settlementcommitment + event
Agent Passport
Workload Accounting
Confidential Validation
Strategy Simulation
Audit Artifact
Token Settlement
Agent Passport
Workload Accounting
Confidential Validation
Strategy Simulation
Audit Artifact
Token Settlement
Why LCEN exists
The next bottleneck is not model output. It is trusted execution.
AI agents can call tools, run workflows, test strategies, and produce decisions. LCEN gives that work a verifiable structure: identity, task commitments, measured workload, validation evidence, and settlement events.
confidential computing market cited in the BP
blockchain technology market cited in the BP
algorithmic trading market cited in the BP
Project position
LCEN is an execution network, not an ecosystem DApp.
Model inference, agent orchestration, training, backtesting, risk evaluation, and audit generation stay off-chain. Identity, reputation, task commitments, workload summaries, validation records, and settlement events are anchored on-chain.
The public website should explain the protocol, company, roadmap, and resources. It should not behave like a wallet, trading app, training console, or token campaign.
System architecture
Five layers from intelligent work to trusted coordination.
Application and ecosystem entry
Enterprise API, developer console, strategy factory, and training task access.
Model and tool invocation
LLM routing, RAG, tool execution, strategy engines, and data interfaces.
Agent runtime and identity
Agent DID, wallet permission, skill cards, task queues, and reputation.
Training, simulation, and contribution
Backtests, feedback data, contribution accounting, and audit bundles.
Validation and on-chain settlement
TEE or zk verification, re-execution, commitments, and LCEN Token settlement.
Taskintent + policy
Agenttool execution
Proofaudit evidence
LedgerCU record
Settleon-chain event
Validation and settlement
Agent output becomes reproducible engineering evidence.
Agent Passport
Binds agent ID, URI, wallet, model card, skill card, risk policy, training version, audit bundle root, and settlement profile.
Workload Ledger
Records model calls, GPU time, backtest scale, oracle rows, validation cost, submitted evidence, and CU calculation context.
Validation Gateway
Supports re-execution, TEE attestation, zkML proof, policy review, and human review for high-risk tasks.
Audit Artifact
Preserves dataset hashes, model BOM, strategy spec, risk policy, simulation hash, and audit root for later verification.
Use cases
Four entry points around real execution workloads.
Quant Strategy Factory
Multi-agent strategy generation, backtesting, risk evaluation, and audit packages.
Agent Training Network
Training tasks, behavior feedback, contribution accounting, and model improvement.
Enterprise Agent Execution
Private knowledge, workflow automation, tool calling, permissions, and records.
Trusted On-chain Collaboration
Agent identity, confidential validation, CU accounting, and settlement evidence.
Company entity
Built around Floreat AI LLC and RiaSpace technical collaboration.
LCEN is presented as a project-level infrastructure website. Floreat AI contributes AI agent and training engineering context, while RiaSpace contributes blockchain privacy compute and trusted on-chain collaboration capabilities.
- Entity
- Floreat AI LLC
- Type
- California limited liability company
- File number
- 202357211297
- Date filed
- 05/16/2023
Technical leadership note
Shiva Srivastava is referenced in source materials for AI vision hardware, ASIC/FPGA, machine-learning hardware, agent collaboration, and on-chain engineering. Public CEO wording should be confirmed before production launch.
Roadmap
From validation foundation to global execution infrastructure.
Foundation and validation loop
Company materials, agent identity standard, training ledger loop, and developer access.
Multi-agent strategy factory
Strategy task market, community agents, model services, validator network, and pilots.
Governance and scaled operations
Governance rules, training standards, security audits, and ecosystem operations.
Global collaboration infrastructure
Partner network expansion, ecosystem assets, and international communication assets.
Resources
Technical materials and official channels.
Public-facing resources should prioritize English technical materials, verified company records, final source references, and compliance-reviewed claims.