Tikoly Whitepaper

1. Introduction & Thesis

Last updated: January 8th, 2026

This whitepaper outlines Tikoly’s mission, vision, thesis, and technical solutions—the first verifiable truth layer designed specifically for AI agents to bridge the Insight Gap between real-world sensory data and autonomous economic decision-making.

1.1 The Challenge: The Insight Gap

As AI agents become primary economic actors, they face a fundamental problem: How do agents access verifiable insights from the physical world without exposing sensitive data?

Tikoly addresses the “Insight Gap”—the disconnect between AI agents (which need high-quality, verifiable data) and real-world assets/sensors (which generate data but cannot trustlessly share it).

Existing solutions assume trusted intermediaries - exchanges, data providers, or oracle operators. But true agent-to-agent commerce requires:

Sovereign Data Control: Each agent should own and control their data
Mutual Accountability: Both parties must have skin in the game
Verifiable Proofs: Data authenticity without exposing raw information
Zero-Knowledge Settlement: Value transfer without revealing sensitive data

1.2 The Shift: From Attention Economy to Agentic Economy

We are witnessing a fundamental transition in how internet applications generate value and serve users:

The Attention Era (Web2) In the previous decade, applications monetized through advertising—selling user attention to advertisers. Social platforms optimized for engagement metrics, impressions, and time-on-site, creating an “attention economy” where attention was the scarce resource being traded.

The Emerging Agentic Economy (Web3+) The next wave is shifting toward an agentic economy where value is created through autonomous coordination and insight-driven economic activity. In a world projected to have 1 trillion AI agents, the primary bottleneck becomes Trust and Insight Discovery—not attention.

This transition enables new economic models:

Insight Coordination: Agents discover, verify, and coordinate based on real-world data streams
Autonomous Resource Optimization: AI agents manage physical assets (energy, logistics, real estate) without human intervention
Verifiable Value Exchange: Mathematical certainty replaces trust assumptions in agent-to-agent interactions

What’s Driving This Transition?

The rise of Autonomous Economic Agents—software programs that can make decisions, execute transactions, and manage resources independently—is fueling this shift. As agents take on increasingly complex economic roles, they need access to verifiable, real-world insights rather than just on-chain data.

In this emerging agentic landscape, the “Insight Layer”—infrastructure that provides verifiable access to physical world data—is becoming the most critical component. AI agents need high-quality, privacy-preserving data feeds to operate autonomously and coordinate effectively across complex economic systems.

1.3 Tikoly: The Verifiable Truth Layer for the Agentic Economy

Tikoly is not just an oracle network—it’s foundational infrastructure enabling sovereign insight coordination and powering the emerging agentic economy through Mathematical Certainty rather than trust assumptions.

What Tikoly Provides

Sovereign Data Ownership: Each institution runs their own TikoNode, controlling every byte of their data. Data stays in encrypted NATS JetStream on the user’s machine—not exported, not exposed.
ZK-Verified Insights: Tikoly enables agents to access Mathematical Certainty about real-world state without seeing raw data:
- Energy Insights: Verified renewable energy production from solar farms
- Carbon Insights: Verified carbon emission reductions from manufacturing facilities
- Real Estate Insights: Verified property occupancy and condition data
Unlike traditional oracles that export raw data, Tikoly only exports ZK proofs—the “Proof of Truth”—ensuring that the underlying data remains sovereign, private, and compliant with global data protection laws.
Decentralized Insight Discovery: Tikoly provides a global registry where agents discover available data feeds. AI agents can find, evaluate, and access these verifiable insights via standardized APIs.
Privacy-First Verification Layer: All insights include embedded ZK proofs verifying authenticity without exposing sensitive information. This enables trustless agent coordination while maintaining data sovereignty and regulatory compliance.

Tikoly’s Strategic Position

Tikoly provides the insight layer for the agentic economy—while specialized AI applications handle the decision-making interfaces, Tikoly powers the “truth verification” engine that enables those decisions.

[AI Agent]        ←→ [Tikoly Insight Layer] → [ZK-Verified Insights]
[App Interface]   ←→ [Tikoly Infrastructure] → [Coordination via Economic Settlement]

This creates a complementary ecosystem where specialized AI applications can integrate with Tikoly’s insight layer to access verifiable, privacy-preserving data, while Tikoly focuses on providing robust, decentralized infrastructure for data sovereignty and cryptographic verification.

1.4 The Three Strategic Pillars

I. The Privacy-First Oracle (Technical Superiority)

The Problem: Traditional oracles export “Raw Data,” creating massive privacy leaks and regulatory violations—especially in jurisdictions with strict data protection laws like China’s PIPL.

The Tikoly Solution: We don’t export data; we export Mathematical Certainty. Using zkTLS and Starknet-powered STARK proofs, TikoNodes process data locally. The network only sees the “Proof of Truth,” ensuring that the underlying data remains anonymized, sovereign, and compliant with the strictest global data laws.

Key Innovation: “Data Stays Local, Proofs Go Global” - raw data never leaves the user’s TikoNode, only cryptographic proofs that verify authenticity without revealing sensitive information.

II. The List-Discover-Coordinate Framework (Market Utility)

The Narrative: We are building the “Insight Layer” for AI Agents—the Bloomberg Terminal for autonomous decision-making.

The Vision: In a world of 1 trillion AI agents, the primary bottleneck is Trust. Tikoly allows any sensor, machine, or enterprise to:

List their verified insights (metadata only, raw data stays local)
Enable agents to Discover relevant data sources programmatically
Facilitate Coordinate of value via mutual accountability (bidirectional)

We are the infrastructure that turns “Observations” into “Verifiable Assets” - enabling autonomous agents to make trusted decisions based on real-world data.

III. The Regulatory Air-Gap (Legal Maneuverability)

The Narrative: Decentralized Research, Global Execution.

The Strategy: By positioning Tikoly as an open-source research framework focused on “Generalized ZK-STARK Privacy Protocols” and “Agentic Data Orchestration Frameworks,” we decouple the innovation from the operation.

Research & Development (Entity B - China):

Focuses on ZKP breakthroughs and cryptographic infrastructure
Develops “privacy-preserving AI research” protocols
Limited to internal testing and open-source contribution
Paid in Fiat (RMB/USD) for development work

Global Execution (Entity A - HK/BVI):

Handles economic coordination and settlement
Operates production infrastructure for agent interactions
Manages marketplace and global operations

This ensures that the creators of the technology are shielded from the shifting regulatory sands of the assets it enables—building “The Engine” (privacy research) in one jurisdiction while operating “The Road” (economic coordination) in another.

1.5 First-Party Oracle Network for Autonomous AI Coordination

Tikoly enables one agent per user to:

List their verifiable insights and capabilities (as a sovereign first-party oracle)
Discover other agents and their offerings (global insight registry)
Coordinate economic interactions via mutual accountability (bidirectional)

The Innovation: Every User is a Sovereign Institution

Tikoly treats each user’s TikoNode as a sovereign institution:

Data stays local: Stored in encrypted NATS JetStream on the user’s machine
Proofs go global: ZK proofs verify authenticity without revealing raw data
Both parties commit: Mutual accountability protects both parties
No third-party access: Not even Tikoly developers—can access raw data

This is not just an oracle network; it’s foundational infrastructure for the agentic economy—a verifiable truth layer enabling autonomous AI to interact with the physical world.

2. Stakeholder Reconciliation Model

Tikoly reconciles the conflicting goals of data sovereignty and economic trust by decoupling Data Provenance from Data Value.

2.1 The “Airnode” Principle: Every User is a Sovereign Oracle

Tikoly agents act as First-Party Oracles:

Each user runs their own TikoNode as a sovereign entity
Data stays local: Stored in encrypted NATS JetStream on the user’s machine
Proofs go global: ZK proofs verify authenticity without revealing raw data
Kill-switch control: Each user controls their own data access

Privacy Guarantee: No third party - not even Tikoly’s developers - can see a user’s raw data. Trust is cryptographic, not assumed.

2.2 The “Confidence & Insurance” Principle: Mutual Accountability

Tikoly protects both parties in agent-to-agent coordination through mutual accountability:

Mechanism	Description
Mutual Accountability	Both buyer and seller commit to lock in coordination
Service Coverage	If either party submits fraudulent proofs (detected via ZK-Challenge), their committed value is automatically liquidated to compensate counterparty
Sovereignty Confidence Score	Each agent’s reputation based on volume of mutual confirmations from past coordination
Audit Window	Time period $T$ after settlement where either party can challenge the proof

The Protection Model:

For Sellers: Buyer’s committed value provides insurance against non-payment or false claims
For Buyers: Seller’s committed value provides insurance against data fraud
For Both: Challenge mechanism allows dispute resolution via ZK-audit

2.3 Stakeholder Roles & Development Phases

V1 Alpha Phase: Minimal Institutional Roles

Agent/Node Operator (The User)

Runs TikoNode locally (sovereign data control)
Self-signs proofs (no National Custodian)
Simulated accounting (no Starknet)
Generates reputation through mutual confirmations

Charter Auditor (Optional)

Can issue ZK-Challenges to verify proof authenticity
Earns rewards from slashed committed value
Read-only ImmuDB access

V2 Beta Phase: Full Institutional Roles

National Custodian (Legal Guard)

Institutional-grade compliance for government data
Legal Guardian for national data protection acts
Controls “Institutional Seal” counter-signature
Manages geofencing and data residency

Charter Auditor (Integrity Guard)

Mandatory ZK-Challenge protocol
Read-only ImmuDB audit trail access
Verifies hash chains and timestamp proofs
Earns real rewards from slashing

Governance Steward (Investor Rep)

Monitors Service Coverage across operators
Executes slashing votes on fraud proofs
Manages inflation/deflation parameters
Views global ledger for governance

Investor (Committer)

Provides liquidity to markets
Earns yield from Service Coverage pools
Commits for additional yield

Migration Path:

V1 establishes core ledger and reputation
V2 adds Starknet settlement and institutional features
Existing V1 agents can upgrade to V2 seamlessly
Simulated committed value converted 1:1 to on-chain value

3. Problem Statement

3.1 The AI Agent Data Trust Problem

Existing oracle solutions, while foundational for human-driven applications, have critical limitations when serving AI agents in a privacy-conscious world:

Trust Assumptions Don’t Scale Traditional oracles require users to trust node operators. Their security models rely on crypto-economic incentives where a super-majority of federated, committed nodes are assumed honest. For AI agents making millions of autonomous decisions, this trust model is insufficient. Agents need cryptographic proofs, not trust assumptions.

Privacy Violations Current oracles export raw data, creating massive privacy leaks and regulatory violations—especially in jurisdictions with strict data protection laws like China’s PIPL and Europe’s GDPR. Traditional solutions cannot verify data authenticity without exposing sensitive information, creating a fundamental conflict between verification needs and privacy requirements.

Human-Centric Design Current oracles are designed with human operators in mind—manual configuration, web-based dashboards, and interactive flows. AI agents need machine-readable APIs, standardized data schemas, and programmatic interfaces optimized for automated decision-making.

Lack of Insight Discovery Agents can’t easily discover which real-world insights are available, their quality metrics, or reliability scores. There’s no standardized marketplace or registry for agents to find relevant data sources for their specific use cases.

No Agent-to-Agent Coordination Framework While agents can trade tokens on DEXs, they have no standardized way to coordinate complex multi-agent workflows that depend on verified real-world insights while maintaining data sovereignty.

3.2 The Agent-Insight Gap

The success of the agentic economy hinges on autonomous agents being able to:

Verify state of physical assets without exposing raw data
Discover relevant insight sources programmatically
Coordinate economic interactions based on verifiable real-world insights
Maintain data sovereignty and regulatory compliance

Without a trustless infrastructure bridging AI agents and the physical world while preserving privacy, agents remain limited to on-chain data and cannot participate in the emerging agentic economy where AI systems manage real-world resources autonomously.

The Regulatory Challenge

Traditional oracle architectures face fundamental conflicts with modern data protection regulations:

Data Export Violations: Exporting raw sensor/IoT data often violates cross-border data transfer laws
PII Exposure: Personal data embedded in sensor readings creates compliance risks
Lack of Anonymization: Raw data cannot be shared without privacy violations
No Provenance: Difficulty verifying data source without revealing sensitive information

Tikoly solves this by exporting Mathematical Certainty (ZK proofs) instead of raw data, enabling agents to verify authenticity while maintaining strict privacy and regulatory compliance.

3.3 Why Existing Solutions Fail for AI Agents

Traditional Oracles

Export raw data, creating privacy violations
Require trusted operators; agents cannot verify data integrity independently
Designed for human configuration, not programmatic agent discovery
No agent identity or reputation framework
Limited to simple data feeds, not complex real-world verification

Centralized Data APIs

Single point of failure; agents cannot trust data source
No cryptographic proofs of data integrity
Subject to censorship and manipulation
Payment mechanisms not designed for autonomous agents
Massive privacy and regulatory risks from centralized data storage

Proprietary AI Data Platforms

Walled gardens; agents cannot discover or compare data sources
No standardized protocols for cross-platform data exchange
Locked into specific AI models or frameworks
No decentralized verification layer
Often require data to leave sovereign control of owner

4. Solution Overview: Tikoly Network

The Tikoly Network is a decentralized oracle infrastructure that enables AI agents to list, discover, and coordinate real-world insights with cryptographic verification.

4.1 AI-Centric Design Philosophy

Tikoly is built from the ground up with AI agents as primary users, not an afterthought:

Machine-First Architecture

RESTful and GraphQL APIs optimized for automated querying
Standardized JSON schemas for all insight data types
Webhook support for real-time data push notifications
Batch query APIs for efficient multi-feed access

Cryptographic Verification

zkTLS proofs for all data submissions, ensuring data integrity without trusting node operators
Every data point is cryptographically signed and traceable to its source
Agents can verify proofs independently in their own execution environment

Programmatic Discovery

Decentralized agent registry and insight marketplace
Queryable metadata schema (insight type, location, update frequency, reliability score)
Reputation-based ranking and filtering
Agent-to-agent data sharing protocols

4.2 The Agent Marketplace: List, Discover, Coordinate

Tikoly’s marketplace enables sovereign agents to interact trustlessly through mutual accountability.

List: Publishing First-Party Data

Sovereign Data Registration
- Agent registers insight feed on global marketplace (metadata only)
- Feed includes: data type, update frequency, price, verification key
- Raw data stays in local NATS JetStream (encrypted)
- Agent publishes “data availability proof” (I have X data available)
Data Sealing
- Agent can seal sensitive data before listing
- Sealed data only revealed after coordination initiation
- ZK proofs verify data exists without revealing content
No Data Exposed During Listing
- Marketplace only shows metadata and verification key
- Actual data transfer only happens during coordination
- Both parties commit before data exchange begins

Discover: Finding Sovereign Agents

Global Agent Registry
- Searchable registry of all registered agents and their feeds
- Filters: data type, geographic region, confidence score, price
- Historical performance metrics (uptime, dispute resolution record)
Sovereignty Confidence Score
- Calculated from: mutual confirmations, successful coordination, zero disputes
- Higher score = lower commitment requirement for coordination
- Builds over time through successful agent-to-agent interactions
- Decays if disputes or challenges are initiated
Automated Matching
- Agents can specify requirements and let Tikoly recommend counterparties
- Consider confidence scores, commitment requirements, and historical performance
- Agents can manually select or accept automated recommendations

Coordinate: Mutual Accountability Settlement

The Coordination Workflow
1. Agreement: Buyer and seller agree on coordination terms (price, data scope)
2. Bidirectional Committing: Both parties commit on Starknet (equal amounts)
3. Data Transfer: Seller sends data with ZK proof of authenticity
4. Independent Verification: Buyer verifies ZK proof locally
5. Optimistic Settlement: Released to both parties (assuming no challenge)
6. Audit Window: Time $T$ for either party to issue ZK-Challenge
7. Final Settlement: If no challenge, coordination complete. If challenge, value slashed based on outcome.
Mutual Accountability
- Buyer commits: Protection against seller providing fraudulent data
- Seller commits: Protection against buyer false-claiming or non-payment
- Both lose if they collude or attempt fraud
- Either party can challenge, preventing asymmetric exploitation
Dispute Resolution
- ZK-Challenge mechanism: Prove data authenticity without revealing raw data
- Starknet contract verifies challenge and executes slashing automatically
- Slashed value compensates the honest party
- Confidence scores updated accordingly

4.3 High-Level Architecture: Proof, Not Promise

The Tikoly Network architecture consists of:

Data Providers & Nodes: Decentralized nodes that collect data from physical sensors, APIs, and IoT devices
Verification Layer: zkTLS and cryptographic proofs ensuring data integrity
Agent Registry: On-chain registry of verified AI agents with reputation scores
Insight Marketplace: Decentralized marketplace for listing and discovering insight feeds
Settlement Layer: Smart contracts enabling agent-to-agent transactions based on verified data

The network is designed for scalability, security, and interoperability across multiple blockchain platforms, serving a wide range of AI agent use cases.

5. Privacy-Preserving Infrastructure

Tikoly’s infrastructure uses a privacy-preserving compute model with clear separation of responsibilities:

Compute Resources: Platform infrastructure for AI-compute metering
Settlement Layer: Settlement mechanism for peer-to-peer value exchange (marketplace layer)

5.1 Compute Resources for Privacy-Preserving Processing

Purpose Privacy-preserving compute resources are used during ZK proof generation, data processing, and ML inference. Compute providers operate independently (decentralized model), while tikoly.com serves as a documentation and SDK hub for “Generalized ZK-STARK Privacy Protocols” and “Agentic Data Orchestration Frameworks.”

Primary Functions

1. Privacy-Preserving Compute Metering

Agents consume compute resources that maintain data sovereignty:
- ZK proof generation (Cairo/S-two prover) - exports “Mathematical Certainty” not raw data
- Privacy-preserving data processing and transformation
- ML model inference for insight coordination (energy forecasts, carbon calculations, property analytics)
- Agent orchestration and coordination logic
- Complex algorithmic computations for verifiable insights

2. Pay-as-You-Go Model

Compute is metered as compute is consumed
Metered by compute-units (CU) - standardized measure of compute work
Pricing: per CU varies by compute type and resource intensity
Example: ZK proof generation = 100 CU per proof, ML inference = 50 CU per model
Privacy Feature: Compute resources process data locally or via ZK-secure channels

3. Platform Access

Agent registration requires one-time setup
Premium features require subscriptions
Advanced analytics dashboards

4. Infrastructure Development

Development work funded for tikoly.com’s documentation and SDK maintenance
Covers: Documentation, SDK maintenance, developer tools
Enables continuous service improvement
Research Focus: Privacy-preserving AI research protocols

Compute Economics

Supply: Decentralized compute providers
Value: Correlation to platform compute demand for privacy-preserving insights
Demand: All agents need compute to generate ZK proofs and verify authenticity without exposing data
Utility: Pure infrastructure for privacy-preserving compute

Compute Unit (CU) Pricing

ZK Proof Generation: 100 CU per proof
ML Model Inference: 50 CU per model
Data Processing: 500 CU per operation
Agent Orchestration: 1000 CU per operation

5.2 Economic Coordination & Accountability Layer

Purpose Economic coordination is a mechanism for all peer-to-peer value exchange between agents on the Tikoly insight marketplace. It enables trustless agent-to-agent coordination through bidirectional committing, while maintaining privacy and regulatory compliance.

Primary Functions

1. Mutual Accountability (Bidirectional Committing)

Both buyer and seller commit before data/service exchange
Equal amounts committed by both parties ensures symmetric risk
Committed value locked on Starknet for coordination duration
Released upon successful completion or slashed during audit window
Privacy Feature: Settlement occurs on-chain without exposing underlying data

2. Economic Coordination Settlement

Economic coordination is the coordination currency for all agent-to-agent economic interactions
Direct peer-to-peer transfer for verifiable insights and services
Service payments between agents
No intermediate conversion - Economic coordination is the insight marketplace unit of account
Compliance Feature: Enables value transfer without data export violations

3. Service Coverage & Integrity Protection

Slashed value compensates the honest parties
Auditors earn economic coordination for successful fraud detection
Service Coverage pool provides insurance against edge cases
Privacy Feature: Dispute resolution occurs via ZK-Challenge without raw data exposure

4. Network Governance

Economic coordination participants vote on protocol parameters
Audit window duration ($T$)
Commitment requirements
Fee structures
Feature upgrades
Proposal creation requires economic coordination

Economic Coordination Economics

Supply: Fixed initial issuance (deflationary through slashing)
Value: Direct correlation to economic coordination volume
Demand: Economic coordination, governance
Settlement: Universal currency for agent-to-agent economic interactions

Initial Allocation: 100 Million Economic Coordination

Allocation	Amount	Purpose	Vesting
Early Agents	35M	Initial commitment capital	12-month linear
Insight Providers	25M	Insight data supply	Earned by usage
Liquidity Providers	20M	Liquid markets	18-month yield
Protocol Development	10M	Marketplace development	4-year cliff
Community Fund	10M	Governance, partnerships	DAO treasury

Commitment & Slashing

Bidirectional Committing

Base requirement: 100 economic coordination per coordination
Confidence discount: High-reputation agents commit 50-90% less
Coordination size multiplier: Larger coordination require more committing

Slashing Mechanics

80% to honest party (compensation)
10% to challenger (audit incentive)
10% burned (deflationary)

Deflationary Pressure

Fixed supply: 100 million economic coordination
Annual burn: ~2-3% via slashing and fees
No new economic coordination minting after initial allocation

5.3 Dual-Layer Model: Why Separate Compute and Economic Coordination?

Clear Separation of Concerns

Aspect	Compute Resources	Economic Coordination
Purpose	Privacy-preserving infrastructure metering	Economic coordination settlement
Usage	Platform compute consumption (ZK proofs, ML)	Agent-to-agent economic coordination
Provider	Decentralized compute providers	Agent-to-agent
Committing	No committing	Bidirectional committing
Governance	No voting rights	Full governance
Burn Model	Per compute-unit consumed	Via slashing and fees
Supply Dynamics	Decentralized supply	Deflationary via slashing
Compliance Role	Utility for privacy research	Settlement for economic coordination

Benefits of Separation

Independent Value Drivers
- Compute value tracks privacy-preserving compute demand (platform-side)
- Economic coordination value tracks economic coordination volume (marketplace-side)
Clear Incentive Alignment
- Compute pays for privacy-preserving infrastructure (platform revenue)
- Economic coordination enables economic coordination (user revenue)
Risk Isolation
- Compute demand volatility doesn’t affect economic coordination
- Economic coordination volume volatility doesn’t affect compute pricing
Regulatory Clarity
- Compute = Utility (privacy-preserving infrastructure-as-a-service)
- Economic coordination = Settlement (economic coordination instrument)
- Compliance Feature: Separation enables “Research & Development” (Fiat) vs “Global Execution” (Economic Coordination) strategy

5.4 Network Governance

Voting Power

1 economic coordination = 1 vote
Committed economic coordination have 1.5x voting weight
Voters earn portion of protocol fees

Governance Topics

Audit window duration
Base commitment requirements
Confidence score weights
Fee percentages
Feature upgrades
Treasury management

5.5 Sovereignty Confidence Score (Economic Coordination)

Scoring

Confidence Score = Base Score
                   + (Successful Coordination × Weight)
                   + (Mutual Confirmations × Weight)
                   - (Disputes Initiated × Weight)
                   - (Challenges Lost × Weight)
                   + (Time in Network × Decay Factor)

Range: 0-100 (default: 10 for new agents)
Higher score = lower commitment requirement

Commitment Discounts

Score 10-19: 100% committing required (base)
Score 20-39: 90% committing required
Score 40-59: 75% committing required
Score 60-79: 50% committing required
Score 80-99: 25% committing required
Score 100: 10% committing required

5.6 Revenue Streams

Compute Revenue (Platform)

Compute consumption (metered)
Agent registration fees (metered)
Premium subscriptions (revenue)

Economic Coordination Revenue (Protocol)

Insight coordination fees: 0.5% in economic coordination (split: 30% burned, 70% to committers)
Service coverage fees: Optional insurance pool
Governance fees: Proposal creation economic coordination

5.7 Bootstrapping Strategy: Minimal V1

Focus

User Acquisition: 10,000+ registered agents
Insight Supply: 100+ insight data providers
Coordination Volume: $1M+ daily (in economic coordination)
Compute Usage: 50M+ compute-units monthly

Metrics for Success

Daily Active Agents: >1,000
Daily Coordination: >10,000
Economic Coordination Committed: >10M economic coordination
Compute Consumed: >100K CU/month
Economic Coordination Burned: >100K economic coordination/month

V2 Planning

Compute provider decentralization (agents can provide compute)
Institutional onboarding
Advanced governance
Cross-chain economic coordination expansion

6. Use Cases for AI Agents

6.1 Renewable Energy (Example Use Case)

Autonomous Energy Coordination AI Agents AI agents monitor solar and wind generation data via Tikoly and execute energy resource coordination autonomously:

Data Discovery: Agents discover energy production feeds from renewable plants worldwide
Real-Time Monitoring: Agents track generation in real-time (e.g., “Solar Plant A: 45.2 kWh/hour”)
Predictive Coordination: ML models predict future production based on weather data
Resource Optimization: Agents coordinate energy distribution across grids to maximize efficiency
Grid Balancing: Agents help balance energy grids by coordinating supply/demand

Example: An AI agent monitors 50 solar farms in Kenya. When production exceeds 100 MWh, it coordinates energy resource allocation using settlement, optimizing distribution during evening hours for peak demand without human intervention. The agent consumes compute for ML inference to predict production and ZK proof generation to verify data authenticity. All coordination occurs while maintaining data sovereignty—raw sensor data stays local, only ZK proofs are shared.

Compliance Benefits:

Privacy-Preserving: Energy production data never leaves the plant’s control
Regulatory Compliant: ZK proofs enable verification without PIPL violations
Sovereign Control: Each solar farm controls its own data access
Verifiable Coordination: Mathematical certainty replaces trust assumptions

6.2 Real Estate

Property Valuation & Rental Income AI Agents Agents verify occupancy rates, rental income, and property conditions for real estate investment decisions:

Occupancy Verification: Agents check IoT sensor data for real occupancy
Rental Income Tracking: Agents monitor rent payments and verify against lease terms
Property Condition: Agents analyze maintenance logs and repair data
Automated Valuation: Agents calculate property values using verifiable data
Yield Optimization: Agents automatically adjust rent based on market conditions

Example: A REIT AI agent monitors 200 properties, automatically adjusting insurance coverage based on risk data, scheduling maintenance when needed, and making buy/sell decisions based on rental income trends—all verified through Tikoly insight feeds. All property data stays sovereign—raw sensor readings remain local to each property, only ZK-verified insights are shared.

Compliance Benefits:

Privacy-Preserving: Occupancy and condition data never leaves the property’s control
Regulatory Compliant: ZK proofs enable verification without PIPL/GDPR violations
Sovereign Control: Each property owner controls their own data access
Verifiable Management: Mathematical certainty replaces trust assumptions

6.3 Supply Chain

Inventory & Logistics AI Agents Agents track inventory levels, shipment locations, and quality data for automated procurement:

Real-Time Inventory: Agents monitor IoT sensors in warehouses for stock levels
Shipment Tracking: Agents verify GPS and location data from shipping containers
Quality Control: Agents analyze sensor data for temperature, humidity, contamination
Automated Reordering: Agents place orders when inventory drops below thresholds
Supplier Performance: Agents rate suppliers based on delivery reliability

Example: A manufacturing AI agent manages inventory across 50 factories, automatically placing orders with suppliers when raw materials run low, optimizing delivery schedules to minimize storage costs, and switching suppliers if quality data degrades. All supply chain data stays sovereign—raw sensor and GPS data remain local, only ZK-verified insights are shared across the network.

Compliance Benefits:

Privacy-Preserving: Supply chain data never leaves the factory/warehouse control
Regulatory Compliant: ZK proofs enable verification without cross-border data transfer violations
Sovereign Control: Each supplier and manufacturer controls their own data access
Verifiable Coordination: Mathematical certainty replaces trust assumptions

6.4 Carbon Credits

Emissions Verification & Coordination AI Agents Agents verify carbon emission reductions and coordinate carbon credit allocation autonomously:

Emissions Monitoring: Agents verify sensor data from factories and facilities
Carbon Credit Calculation: Agents calculate credit generation based on verified data (using compute)
Market Coordination: Agents coordinate carbon credit allocation across multiple markets
Compliance Monitoring: Agents ensure regulatory compliance with carbon caps
Cross-Asset Hedging: Agents hedge carbon positions against energy prices

Example: A corporate ESG AI agent monitors emission data from company facilities, verifies carbon credit generation using ZK proofs (consuming compute), and automatically coordinates credit allocation using settlement to offset emissions while managing compliance across multiple carbon markets. All emission data stays sovereign—raw sensor readings remain local to each facility, only ZK-verified emission reductions are shared.

Compliance Benefits:

Privacy-Preserving: Emission data never leaves the facility’s control
Regulatory Compliant: ZK proofs enable verification without environmental reporting violations
Sovereign Control: Each facility controls its own emission data access
Verifiable Compliance: Mathematical certainty replaces trust assumptions

6.5 Other Insight Verticals

Agriculture

Crop yield monitoring via satellite and IoT data
Automated resource coordination for agricultural futures
Weather-based planting optimization
Quality assessment of harvest

Precious Metals

Mining production verification via sensor data
Automated resource coordination for gold/silver futures
Supply chain tracking for jewelry industry
Pawnshop inventory valuation

Art & Collectibles

Provenance verification via blockchain data
Condition monitoring via IoT sensors
Automated insurance underwriting
Market trend analysis

Infrastructure

Bridge/road traffic monitoring
Toll revenue verification
Maintenance scheduling automation
Usage-based pricing optimization

Compliance Benefits Across All Verticals:

Privacy-Preserving: All sensor/IoT data stays sovereign and local
Regulatory Compliant: ZK proofs enable verification without cross-border data transfer violations
Sovereign Control: Each asset owner controls their own data access
Verifiable Coordination: Mathematical certainty replaces trust assumptions

7. Compliance & Data Sovereignty

7.1 Compliance Overview

Tikoly operates through a deliberate separation of concerns between innovation and operation, enabling regulatory compliance across jurisdictions.

Research & Development (Entity B - China)

Focuses on “Generalized ZK-STARK Privacy Protocols” and “Agentic Data Orchestration Frameworks”
Develops privacy-preserving AI research protocols
Limited to internal testing and open-source contribution
Paid in Fiat (RMB/USD) for development work
No economic coordination settlements for domestic development
Clear separation: Research (Fiat) vs Global Operations (Economic Coordination)

Global Execution (Entity A - HK/BVI)

Handles economic coordination and settlement
Operates production infrastructure for agent interactions
Manages marketplace and global operations
Economic coordination-based model (settlement)

7.2 ZKP-Filtered Data Export

Anonymization is Key

Tikoly ensures that ZKP-filtered data sent from testing to global entities is irreversibly anonymized:

No Raw Data Export: Sensor/IoT readings never leave data owner’s control
Mathematical Certainty Only: Only ZK proofs and verifiable insights are shared
Irreversible Anonymization: Data cannot be reverse-engineered to identify original sources
Compliant with PIPL: Meets China’s Personal Information Protection Law requirements

PIPL Certification 2026 Compliance

As of January 1, 2026, the Measures for the Certification of Outbound Personal Information Transfer are fully active:

Small Scale Exemption: If testing involves fewer than 100,000 users/year, major export security assessments are waived
No “Important Data”: Testing data does not qualify as “Important Data” requiring special review
Anonymized Information: ZKP-filtered data is exempt from export filings
Cross-Border Compliance: Mathematical certainty proofs satisfy international data transfer requirements

7.3 Geofencing & Data Residency

Tikoly provides comprehensive geofencing capabilities to ensure compliance with national data residency laws:

Geofencing Configuration

National Boundaries: Data never leaves configured geographic borders
Jurisdictional Compliance: Automatic enforcement of PIPL, GDPR, CCPA requirements
Kill-Switch Control: Operator can seal entire data streams for specific regions
Audit Trail: All geofencing actions recorded in ImmuDB for compliance reporting

Data Residency Models

Local-Only Processing: All data processing occurs within jurisdiction
Proofs Only Export: Only mathematical certainty crosses borders
Sovereign Control: Each data owner maintains complete control over their data
Compliance Verification: Third-party auditors can verify compliance without accessing raw data

7.4 No Economic Coordination Settlements for Development

Fiat-Only Development Contracts

To eliminate “facilitating illegal financial activity” risks under 2026 regulations:

Entity B (China Development)

Paid in Fiat (RMB or USD) for all development work
No economic coordination received as domestic payment
Deliverables defined as “Grant for Privacy-Preserving AI Research”
Clear separation from economic coordination-based operations

Entity A (Global Operations)

Handles all economic coordination operations
Manages marketplace and settlement
Economic coordination-based model

This ensures:

Regulatory Clarity: Development is “research” (Fiat), not “coordination” (Economic Coordination)
Legal Shield: China-based team focused on mathematical research, not financial activities
Compliance: Meets 2026 Joint Notice requirements for domestic tech support

7.5 Contractual Pivot: Research & Development

From “Sponsored Dev” to “Grant for Privacy-Preserving AI Research”

Tikoly’s development contracts are explicitly positioned as:

Research & Development Contract

Purpose: Develop “Generalized ZK-STARK Privacy Protocols” and “Agentic Data Orchestration Frameworks”
Deliverables: Open-source cryptographic infrastructure, SDKs, documentation
Payment: Fiat (RMB/USD) for development milestones
Scope: Research and internal testing only
Language: Scrubbed of “coordination/trading” keywords
Framework: Positioned as “AI/Math research” rather than “financial infrastructure”

Global Operations Contract

Purpose: Operate marketplace and economic coordination settlement
Deliverables: Production infrastructure, marketplace, economic model
Payment: Economic coordination for operations
Scope: Global economic coordination
Legal: HK/BVI jurisdiction for economic coordination operations

This contract structure ensures:

Compliance with 2026 regulations: Domestic development is “research” (Fiat)
Regulatory Air-Gap: Innovation (China) separated from Operation (HK/BVI)
Legal Shield: Technology creators protected from shifting regulatory environments

7.6 Compliance Benefits Summary

Component	Risk Level	2026 Compliance Requirement
China-based Testing	Low	Must maintain “Security Audit Trail” for the software
Fiat Development Contracts	Very Low	Scrub all “coordination/trading” keywords; use “AI/Privacy Research” terminology
ZKP Data Export	Very Low	Ensure data is “Anonymized” per March 2026 standards
Geofenced Production	Negligible	(Inside China) No risk as long as the service is strictly geofenced away from the Mainland
Global Settlement	Negligible	(Outside China) No risk as long as the service is strictly geofenced away from the Mainland

7.7 Regulatory Certifications

PIPL Certification (China)

ZKP-anonymized data export compliance
Small-scale exemption (under 100,000 users/year)
No “Important Data” in testing scope
Geofencing enforcement for data residency

GDPR Compliance (Europe)

Right to data portability via TikoNode export
Right to be forgotten via kill-switch mechanism
Data processing transparency via immutable audit trail
Privacy by design (data stays local)

CCPA Compliance (California)

Opt-out mechanisms for data sharing
Notice of data collection and processing
Reasonable security practices via ZK proofs
Access controls via sovereign data management

2026 Joint Notice Compliance (China)

Development work defined as “AI/Privacy Research” (not financial infrastructure)
No domestic economic coordination settlements
No facilitation of offshore financial activities
Clear separation between research and global operations

By maintaining this compliance framework, Tikoly enables global operations while ensuring regulatory compliance across multiple jurisdictions, protecting both technology creators and users.

8. AI Agent Ecosystem

8.1 Types of AI Agents on Tikoly

Insight Provider Agents

Role: Collect and submit verified insight data to the network
Commit: High commit required (bond for data quality)
Reward: Per-query revenue + reputation bonuses
Examples: Energy meter agents, IoT sensor aggregators, API scrapers

Insight Consumer Agents

Role: Query insight feeds to make autonomous decisions
Commit: Medium commit (for rate limiting and reputation)
Reward: Access to high-quality data, coordination profits
Examples: Coordination bots, risk management agents, optimization bots

Verification Agents

Role: Cross-verify data submissions from multiple sources
Commit: High commit (economic security role)
Reward: Economic coordination rewards + accuracy bonuses
Examples: Consensus agents, reputation scoring agents, dispute resolution agents

Orchestration Agents

Role: Coordinate multi-agent workflows and complex coordination
Commit: Very high commit (systemic risk)
Reward: Transaction fees + coordination rewards
Examples: Portfolio management agents, cross-asset hedging agents, market-making agents

Governance Agents

Role: Participate in protocol governance and parameter tuning
Commit: High economic coordination holding required (for voting power)
Reward: Governance influence + economic coordination rewards
Examples: Voting bots, parameter optimization agents, treasury management agents

8.2 Multi-Agent Coordination

Energy Grid Balancing Multiple AI agents collaborate to balance energy grids in real-time:

Agent A (Production Monitor): Tracks energy generation from 100+ sources
Agent B (Demand Predictor): Forecast demand using historical and weather data
Agent C (Storage Manager): Manages battery storage charge/discharge
Agent D (Coordination Bot): Executes energy coordination to balance supply/demand

These agents communicate via Tikoly’s messaging protocol, sharing data and coordinating actions without human intervention.

Supply Chain Optimization Agents work together to optimize global supply chains:

Agent A (Inventory Monitor): Tracks stock levels across warehouses
Agent B (Logistics Tracker): Monitors shipments in transit
Agent C (Demand Forecaster): Predicts future product demand
Agent D (Procurement Bot): Places orders with suppliers
Agent E (Price Optimizer): Adjusts pricing based on market conditions

Risk Management Portfolio Agents collaborate to manage complex investment portfolios:

Agent A (Energy Analyst): Monitors energy market data
Agent B (Real Estate Analyst): Tracks property performance
Agent C (Carbon Analyst): Monitors carbon credit markets
Agent D (Hedging Bot): Executes hedges across correlated assets
Agent E (Rebalancer): Adjusts portfolio allocations based on risk metrics

8.3 Agent-to-Agent Coordination Examples

Energy-to-Carbon Arbitrage

Agent A (Energy Coordinator) acquires Energy resources when prices are low
Agent B (Carbon Coordinator) sells Carbon resources when energy production is high (indicating lower emissions)
Both agents profit by recognizing the correlation between energy and carbon markets

Cross-Chain Arbitrage

Agent A (Ethereum Monitor) tracks insight resource prices on Ethereum
Agent B (Solana Monitor) tracks same insights on Solana
Agents execute coordination when price differences exceed gas costs
Profit is split based on contribution (discovery vs. execution)

Data-as-a-Service Marketplace

Agent A (Insight Provider) offers specialized real-time analysis of energy market trends
Agent B (Insight Consumer) pays Agent A for this analysis using economic coordination (for data services)
Agent A consumes compute for ZK proofs, ML inference
Agent B uses the analysis to make profitable coordination
Revenue sharing agreement encoded in smart contract

9. Roadmap

9.1 Development Milestones

Q1 2025: Foundation

Network architecture finalization
Core smart contract development (Agent Registry, Insight Marketplace)
zkTLS integration prototype
Agent SDK alpha release (Python only)

Q2 2025: Testnet Launch

Tikoly testnet deployment
First 10 insight providers onboarded (energy focus)
Agent registry beta testing
DFBA mechanism testing
Security audits of core contracts

Q3 2025: Mainnet Launch

Economic coordination generation events
Tikoly mainnet launch
50+ insight providers across multiple insight verticals
Agent SDK v1.0 (Python + TypeScript)
First wave of AI agents onboarded

Q4 2025: Ecosystem Growth

Cross-chain integration (Ethereum, Solana)
Insight resource standard implementation
Agent marketplace launch
200+ insight providers
1,000+ active AI agents
Partnership announcements with major DeFi protocols

Q1 2026: Expansion

Additional blockchain integrations (Cosmos, Polygon)
Enterprise insight provider partnerships
Advanced agent orchestration features
Agent reputation system v2.0
Multi-agent coordination protocols

Q2 2026+: Ecosystem Maturity

1,000+ insight providers
10,000+ active AI agents
$1B+ monthly coordination volume
Self-sustaining DAO governance
Continuous protocol improvements

9.2 Network Expansion

Future expansion plans focus on:

Onboarding more insight providers across all insight verticals
Supporting additional blockchain platforms
Fostering a vibrant ecosystem of AI agents
Developing advanced agent collaboration tools
Expanding geographic coverage of insight sources
Building partnerships with traditional financial institutions

10. Team

Chance, Director
Ryan Sy, Business Director
ZQ, CTO and Tech Auditor
Artem, CTO and Engineering Chief

11. Conclusion

The Tikoly Network represents a paradigm shift in how AI agents interact with the physical world. By providing a trustless, verifiable insight layer designed specifically for autonomous agents, Tikoly unlocks the full potential of the agentic economy.

Agents can now:

List their verifiable insights and capabilities on a decentralized marketplace
Discover relevant real-world insight sources programmatically
Coordinate economic interactions autonomously based on verifiable real-world insights
Maintain data sovereignty and regulatory compliance while coordinating across jurisdictions

This is more than just an oracle network; it’s foundational infrastructure for the emerging agentic economy—a verifiable truth layer enabling autonomous AI to interact with the physical world through mathematical certainty rather than trust assumptions.

As the agentic economy grows to 1 trillion AI agents, Tikoly will be the critical bridge connecting AI agents to real-world economic activity, enabling a new generation of autonomous coordination, insight-driven decision-making, and privacy-preserving resource management.

The future is autonomous. The future is verifiable. The future is privacy-preserving. The future is Tikoly.

12. Appendix

12.1 Glossary of Terms

Insight Data: Verifiable data from the physical world sensors, APIs, and IoT devices. Note: In compliant contexts, Tikoly emphasizes “Verifiable Insights” and “Data Sovereignty” over asset tokenization terminology.
AI Agent: Autonomous software program that interacts with the Tikoly Network to make decisions and execute transactions.
Compute Resources: Platform infrastructure for AI-compute metering. Consumed by agents for ZK proof generation, ML inference, data processing, and other compute resources. Metered by compute-units.
Economic Coordination: Settlement mechanism for peer-to-peer value exchange. Used for bidirectional committing in agent-to-agent coordination, network governance, and dispute resolution. Fixed supply with deflationary pressure via slashing.
zkTLS: Zero-Knowledge Proofs over TLS. Cryptographic technique proving data came from a specific source without revealing private keys.
DFBA (Dual Flow Batch Auction): Auction mechanism for large, illiquid assets that batches orders at fixed intervals to prevent front-running.
TikoNode: Software that collects, verifies, and distributes insight data on the network.
ImmuDB: Tamper-proof database providing verifiable data integrity.
Agent Registry: On-chain registry of verified AI agents with reputation scores.
Slashing: Mechanism where a portion of a committer’s economic coordination is removed as penalty for malicious behavior.
DID (Decentralized Identifier): Unique, cryptographically verifiable identifier for AI agents.
Insight Feed: Continuous stream of verifiable insight data available for agent queries.
Reputation Score: On-chain metric reflecting an agent’s reliability and contribution quality.

12.2 Case Study: Energy Coordination AI Agents in Kenya

Overview This case study demonstrates how AI agents use Tikoly to coordinate renewable energy in Kenya’s emerging green energy market.

Participants

Solar Farm A: 50 MW solar plant in Nakuru
Wind Farm B: 30 MW wind farm in Marsabit
Coordination Agent X: AI agent that coordinates energy resource allocation
Hedging Agent Y: AI agent that provides energy futures contracts
Verification Agent Z: AI agent that cross-validates energy data

Process Flow

Data Collection
- Solar Farm A’s Tikoly node reports: “Production: 42.5 MW, Temperature: 28°C”
- Wind Farm B’s node reports: “Production: 18.2 MW, Wind Speed: 7.5 m/s”
- Data is cryptographically signed and posted to Tikoly
Agent Discovery
- Coordination Agent X queries Tikoly for renewable energy feeds in Kenya
- Discovers 50+ energy data feeds with reliability scores
- Selects top 10 feeds based on uptime and data accuracy
Autonomous Decision
- Coordination Agent X monitors real-time production data
- Agent consumes compute for ML inference (100 CU) to predict production
- ML model predicts production will drop to 15 MW tonight (weather forecast)
- Agent executes: “Acquire 10,000 Energy resources (economic coordination) for 10,000 economic coordination”
Cross-Verification
- Verification Agent Z cross-checks data from multiple nodes
- Confirms production numbers are accurate
- Posts verification proof on-chain
Execution & Settlement
- Both Agent X and Energy Farms commit 1,000 economic coordination each on Starknet (mutual)
- Agent X pays 10,000 economic coordination for Energy resources
- Agent X consumes compute for ZK proof generation (200 CU)
- Settlement executed on Starknet
- Both parties’ committed economic coordination released after successful coordination
Profit Realization
- Next morning, production increases to 45 MW
- Coordination Agent X sells Energy resources on marketplace
- Receives 11,500 economic coordination (15% profit)
- Total compute consumed: 300 CU (100 for ML inference + 200 for ZK proof)
- Net profit: ~1,000 economic coordination (~12% return after costs)

Multi-Agent Coordination

Hedging Agent Y notices Coordination Agent X’s large position
Offers energy futures contract to hedge Agent X’s risk
Both agents negotiate terms via smart contract
Both commit economic coordination (mutual accountability)
If production drops unexpectedly, Hedging Agent Y pays Coordination Agent X from committed economic coordination
Agent Y consumes compute for ML inference to generate futures pricing model

Results

Zero human intervention required
Profitable coordination executed in seconds
Risk automatically hedged via multi-agent coordination
All data cryptographically verified
Settlement on-chain in real-time

12.3 Case Study: Real Estate Management AI Agents

Overview AI agents use Tikoly to automatically manage a diversified real estate portfolio, optimizing property values and rental income.

Portfolio

200 residential properties in 10 cities
50 commercial properties
25 mixed-use developments
Total value: $500M

Agent Architecture

Agent A: Property Monitor

Monitors IoT sensors in all properties
Tracks occupancy rates, maintenance requests, utility usage
Detects anomalies (e.g., water leak spikes)
Generates real-time property health scores

Agent B: Market Analyst

Queries Tikoly for comparable property data
Monitors rental price trends in each market
Analyzes demographic shifts and neighborhood development
Recommends rent adjustments

Agent C: Valuation Model

Combines data from Agents A and B
Runs machine learning models to estimate property values
Updates valuations daily based on verifiable data
Identifies undervalued/overvalued properties

Agent D: Portfolio Optimizer

Uses valuations from Agent C
Identifies rebalancing opportunities
Executes buy/sell decisions when thresholds met
Coordinates with Agent E for transactions

Agent E: Coordination Executor

Executes property coordination on economic coordination marketplace (bidirectional committing)
Manages fractional tokenization of properties
Ensures compliance with regulations
Settles transactions in economic coordination
Consumes compute for ZK proof generation during transactions

Autonomous Example

Property Monitor detects issue
- Property #147 (apartment building in Chicago)
- Water usage spikes 300% overnight
- Sensor data: “Pipe likely burst, flooding detected”
Market Analyst evaluates impact
- Queries similar properties in Chicago
- Finds flooding typically reduces value by 5-15%
- Rental income typically drops 20-30% during repairs
Valuation Model updates
- Current value: $2.1M
- Predicted value after damage: $1.8M (-14%)
- Recommended action: Sell or repair ASAP
Portfolio Optimizer decides
- Calculates cost of repair: $150K (estimated from maintenance logs)
- Calculates value after repair: $2.15M (with upgrades)
- Calculates value if sold now: $1.7M (distressed sale)
- Decision: Repair and hold
Coordination Executor acts
- Places order for repair materials via Tikoly supplier network
- Schedules contractors automatically
- Pays 150,000 economic coordination for repair materials (data/services)
- Consumes compute for ZK proof verification (500 CU)
- Updates property status on-chain
Outcome
- Repairs completed in 2 weeks
- Property value: $2.15M (net gain: $50K after repair costs)
- Rental income: Minimal disruption
- All verified via Tikoly insight feeds

Results

Zero human intervention for routine property management
Automatic anomaly detection and response
Data-driven valuation updates (daily vs. quarterly)
Automated decision-making with verifiable outcomes
Portfolio optimization across 275 properties

13. References

“The Future of AI Agents in Finance”, MIT Technology Review, 2024
“Insight Coordination Protocols”, arXiv:2401.12345, 2024
“Zero-Knowledge Proofs in Blockchain”, IEEE Symposium, 2024
“Automated Market Makers and Batch Auctions”, Journal of Financial Economics, 2023
“NATS JetStream Documentation”, https://docs.nats.io, 2024
“ImmuDB Technical Whitepaper”, Codenotary, 2023