Agentic Valence
Trust and authorization infrastructure for autonomous commerce — identity, mandates, continuous monitoring, agent security, machine-native commerce, and a ZK trust layer anchored to public blockchains.
Co-Founder • Product Strategy & Fundraising
Human commerce relies on identity, enforceable rules, dispute evidence, monitoring and reputation. Agentic commerce removes the human from the transaction, but not the need for those protections. Agentic Valence is building an open trust layer so agents from different companies can identify themselves, operate inside delegated authority, pay and prove what happened, and build portable trust without exposing unnecessary personal data.
The invisible trust layer behind commerce
- Counterparties can ultimately be identified and held accountable.
- Rules, payment networks, courts, chargebacks and dispute processes constrain bad behavior.
- Reviews and reputation create consequences that follow market participants.
The same questions, at machine speed
- Who is this agent and is there an accountable principal behind it?
- May it take this action or move this amount of value?
- Did it pay or deliver, and can either side prove it later?
- How has it behaved without creating a permanent personal profile?
Five design choices, not one point feature
Seven protocol businesses anchored to Skyrmion
Existing foundation vs. what this round builds
What is already in place
- Founding team spanning enterprise AI/ML, agent architecture, blockchain/ZK and commercialization.
- Defined trust-layer architecture, seven open protocol roles and six coordinated product brands.
- Detailed trust flows, entry-gate logic, business models, buyer maps and customer use cases.
- Corporate, fundraising and diligence structure ready to support investor review.
What $2M is intended to unlock
- Freeze core specifications and advance priority IP filings.
- Launch the initial Skyrmion trust layer with identity and monitoring capabilities.
- Build mandate-controlled payment paths and validate with design partners.
- Deliver private payment proofs and an integrated identity → mandate → payment demonstration.
Distinctive names, immediate functional meaning
Each brand carries a technical identity while the descriptor tells an investor exactly what the product does.
Agentic commerce creates a new trust-and-payments infrastructure problem
Castle Island describes its focus as public-blockchain infrastructure and applications, with investments spanning payments/stablecoins, financial-market infrastructure, cybersecurity and ZK/L2 infrastructure. Wyatt focuses on companies building and expanding the onchain economy.
Public references: Castle Island thesis · Wyatt bio · Portfolio · Payments & Stablecoins
The wedge sits where AI-agent control meets machine-native payments: authority, security, transaction proof and public-chain settlement infrastructure.
Verifiable shared state
Skyrmion is designed to anchor proofs, commitments, registries and stakes without becoming a per-user data store.
Machine-authorized value movement
Mandate + Gluon + payment proofs connect delegated authority to agent negotiation and transaction evidence.
Trust after authentication
Eigen Monitor and Graviton Security add continuous behavior and runtime controls around autonomous actions.
Technical depth + company-building + execution
The team combines enterprise AI architecture, startup/product building, blockchain/ZK direction, and hands-on ML engineering. This is the team Wyatt would be underwriting—not just the protocols.
Viet Le
Original company founder focused on the commercial and blockchain/ZK side of the platform, with responsibility for company development, partnerships, and the trust-layer direction.
- Leads blockchain/ZK strategy and the Skyrmion Network / settlement direction.
- Owns business development, strategic partnerships, commercialization, and company operations alongside the CEO.
- Connects the open-protocol vision to merchant, payment, and ecosystem opportunities.
Shayan Hemmatiyan
AI architect and startup builder with a Ph.D. in computational physics and experience taking multi-agent and GenAI systems from architecture through enterprise deployment.
- Lead AI Architect at Nimble Gravity, leading AI architecture and GenAI initiatives.
- Former co-founder / Principal ML Architect at Newcortex; earlier Principal ML work at LiquidX and senior ML engineering at NCR.
- Leads Agentic Valence product strategy, AI architecture, fundraising, enterprise partnerships, and investor/customer technical discussions.
Khoi Pham
Hands-on ML engineer supporting model development, evaluation, integrations, and product implementation across the agentic trust stack.
- Currently with Knowins and pursuing a Master's in AI at Northeastern University.
- Prior project work includes LLM/eDiscovery research, software engineering, and AI-assistant development.
- Supports evaluation, data/model pipelines, agent integrations, and implementation across the platform.
People use agents; agents use open trust services; Skyrmion anchors the evidence
Negotiation and most service activity stay offchain. Trust providers publish only the roots, proofs and aggregates needed for independent verification. The shared layer can settle to Ethereum while personal records remain offchain.
Attest once, prove later
A user verifies a fact once with an attestation provider. Later the user's agent can prove a derived fact—such as “age ≥ 21”—without revealing the underlying date of birth. The counterparty verifies the proof against the shared root.
Mandate before money moves
An investing or purchasing agent requests a transaction. The payment/custody service asks an independent mandate enforcer for approval. If the request is outside the user's limits, the action is refused; approvals/refusals can leave re-checkable evidence.
Private qualification, objective acceptance
Explore the deeper technical / protocol view
Seven open protocols, independent businesses, one trust layer
Click any box to see why it exists, who uses it, and how it can become a standalone business.
Where the trust layer becomes commercially useful
Examples below illustrate the product logic; they are not claims of current customer deployment.
Age-restricted purchase
An agent proves the buyer satisfies an age threshold without revealing a date of birth or permanent identity profile.
Delegated investing / purchasing
A principal sets hard limits. The enforcer blocks actions outside those limits, while monitoring detects abnormal behavior.
Private payment proof
An agent proves that a qualifying payment occurred for a specific order without exposing more payer information than the counterparty needs.
Merchant compliance
A merchant accepts autonomous orders only after verifying both the agent's authority and any relevant buyer/transaction attributes.
Merchants compete for user intent
A buyer agent publishes a structured need; qualified merchant agents compete directly, reducing dependence on a single marketplace intermediary.
Cross-platform safety and trust
Tool providers detect repeated malicious agent behavior while merchants and buyers carry portable reputation signals across platforms.
Three concrete phases inside the $2M primary plan
The latest product plan makes the sequencing more specific. The current financing strategy remains the $2M primary raise for approximately 12 months.
- Freeze core Identity, Mandate, Watcher and trust-layer specifications.
- Advance initial IP/patent filings around priority trust mechanisms.
- Launch initial Skyrmion environment with first user agent, issuer/attestation provider and watcher path.
- Complete a security baseline review of circuits and key management.
- Build payment rails and independent mandate enforcement.
- Open a test environment with design partners.
- Go live with user-set limits and re-verifiable approval/refusal stamps.
- Start monetization tests around identity verification and mandate enforcement.
- Security review of the payment + mandate path.
- Build and test private payment proofs alongside ordinary payment rails.
- Validate with design partners and move the strongest flow live.
- Expand monetization experiments to proof/verification fees.
- Harden security and deliver an integrated identity + mandate + private-payment demonstration.
Lead with the $2M / 12-month operating plan
The objective is to fund a clear milestone set. Instrument, valuation/cap and ownership are intentionally not fixed in this Pre-NDA package and should be discussed after investor feedback and with counsel.
Strategic bridge if Castle Island prefers to start smaller or anchor a syndicate.
Best balance between technical ambition, validation and disciplined runway.
Broader engineering and security capacity plus multiple pilot/customer environments.
The $2M plan funds product proof, protection and validation
Pre-NDA view shows allocation by category. Detailed compensation, hiring and vendor assumptions are available in confidential diligence.
| 12-month primary plan | Allocation | Purpose |
|---|---|---|
| Core team + employer costs | 37.5% | Keep the founding team fully focused and cover employment burden. |
| Engineering / specialist contractors | 15% | Integration, productization and specialized implementation capacity. |
| Compute / infrastructure | 10% | Local/Blackwell-class compute where justified plus cloud/API capacity. |
| Security / cryptography | 9% | Threat modeling, implementation hardening, cryptographic support and review. |
| Pilots / deployment | 8% | Customer/design-partner integration and controlled pilot environments. |
| IP / legal / compliance | 7.5% | Priority filings, ownership documentation, financing/legal and regulatory work. |
| GTM / conferences / BD | 6.5% | Targeted customer development, ecosystem partnerships and selective events. |
| Operating reserve | 6.5% | Protect milestone execution against timing and vendor uncertainty. |
$500K / ~6 months
Focused first close for core integration, priority IP and investor/customer evidence. Not intended for broad hiring.
$2M / ~12 months
Recommended operating case: integrate, protect, validate and pilot.
$3M / ~18 months
Adds broader engineering capacity, deeper validation and multiple pilot/customer environments.
Seven open protocol services + the shared trust layer
The new working model is more explicit: each protocol can support multiple operators and revenue models. The company intends to ship initial reference implementations and monetize them while preserving interoperability.
| Protocol / product mapping | Likely buyers / operators | What is monetized | Model hypothesis |
|---|---|---|---|
| Identity Isospin IDPrivate Agent Identity | Identity/KYC providers, wallets, agent platforms, enterprises | Agent identity, attributes, attestations and private verification | Per verification + API/licensing + enterprise integration |
| Intent Book Gluon CommerceIntent, Commerce & Reputation | Merchants, marketplaces, discovery platforms, book operators | Order hosting, matching, market listening and completed deals | Success/routing fee + data/listening subscription |
| Mandate Enforcer MandateDelegated Authority & Policy | Users, wallets, fintechs, payment/custody services, agent platforms | Independent policy checks, limits and approval stamps | Subscription + per-transaction/policy-check fee |
| Watcher Eigen MonitorBehavior & Risk Monitoring | Users, merchants, fintech risk teams, agent platforms | Fraud, anomaly and behavior monitoring | Subscription + managed monitoring |
| Guardian Graviton SecurityAgent Runtime Security | Tool/API/MCP operators, enterprise security teams | Runtime/tool protection and cross-server malicious-behavior detection | Security subscription/API + enterprise licensing |
| z402 Skyrmion payment-proof module | Payment providers, APIs, merchants, agent platforms | Private proof that a qualifying payment occurred | Per proof + enterprise/payment integration |
| Reputation Gluon CommerceIntent, Commerce & Reputation | Merchants, buyers, marketplaces, scoring operators | Portable buyer tiers, merchant feedback and score roots | Member/operator fees + protocol/verification fee share |
| Skyrmion Network Shared ZK trust layerZK Trust & Settlement Layer | Protocol operators, fintech/payment infrastructure, ecosystem partners | Proof/commitment anchoring, registries and settlement coordination | Network/service fees + operator integrations |
Identity and ecosystem participants
Identity/KYC/attestation providers, wallet/device platforms and other organizations that can issue or consume private agent credentials.
Users, merchants and payment infrastructure
Consumers/power users, merchants, wallets, fintechs and agent platforms that need bounded autonomous actions and verifiable compliance.
Marketplaces and AI-security teams
Discovery platforms and merchants for the Intent Book; CISOs, tool providers and AI-platform teams for Guardian security.
Open services can reinforce each other
More issuers improve reusable attestations; more mandate/payment activity creates more proof demand; more commerce creates more reputation and monitoring value.
Agentic commerce is moving from assisted decisions toward executed transactions
Platforms are opening agent flows
OpenAI's ACP and Google's UCP are standardizing merchant/catalog/checkout interactions for AI-driven commerce.
Agent-specific rails are arriving
AP2, x402, Mastercard Agent Pay and Visa Trusted Agent Protocol show payments networks actively adapting to autonomous software.
Identity and reputation are moving onchain
ERC-8004 standardizes agent identity, reputation and validation registries—but explicitly treats payments as outside its scope.
Public references: OpenAI ACP · Google UCP · FIDO / AP2 · x402 · Mastercard Agent Pay · Visa TAP · ERC-8004 · McKinsey
The market is active—but fragmented across trust questions
| Category | Representative examples | What they validate | Remaining opening for Agentic Valence |
|---|---|---|---|
| Payment / card-network trust | Mastercard Agent Pay; Visa Trusted Agent Protocol | Agent recognition, authenticated intent, payment authorization and merchant visibility | Cross-rail private proofs, open mandate enforcement, portable reputation and one shared multi-service trust layer |
| Open commerce / payments | OpenAI ACP; Google UCP/AP2; x402 | Merchant connectivity, checkout, authorization and internet-native payments | Privacy-preserving agent identity, independent mandates, continuous monitoring/security and re-checkable cross-platform trust evidence |
| Onchain identity / reputation | ERC-8004 and related registries | Portable identity, reputation and validation signals | Agentic commerce/payment enforcement and privacy-preserving end-to-end trust workflow |
| Enterprise agent security | NHI / bot-defense / AI-security products | Enterprise identities, agent traffic controls and runtime security | Shared cross-company trust proofs connected directly to commerce, mandates and reputation |
Low-friction first review; controlled disclosure for confidential material
This page is the Pre-NDA layer. Detailed corporate, ownership, IP, security, financial and partner materials are kept in a controlled diligence room and released only when appropriate.
Pre-NDA review
Company, team, high-level architecture, buyers, market, roadmap and financing plan.
Confidential diligence
After identity verification, NDA where appropriate, and founder approval.
Controlled access
Named access, logging, expiry/revocation and watermarking where practical.
Shareable in this package
- Investor thesis and Castle Island fit
- Founding team
- High-level product and trust flow
- Commercial hypotheses and likely buyers
- Market/competitive framing
- 6/12/18-month roadmap and financing plan
Hold back until requested
- Current corporate records and cap table
- Invention disclosures, patent drafts and detailed technical moat
- Detailed Skyrmion/ZK, cryptography and security architecture
- Detailed financial model and compensation assumptions
- Partner/pilot evidence subject to disclosure rights
What we expect an institutional investor to evaluate
Corporate & IP
Legal entity, ownership, board approvals, IP assignment and priority invention ownership.
Technology & security
Protocol specifications, integrated demos, Skyrmion architecture, threat model, security boundaries and roadmap.
Commercial evidence
Buyer discovery, pricing hypotheses, partner/design-partner evidence, pilots and GTM strategy.
Team references: Knowins — Who We Are · Shayan LinkedIn · Khoi LinkedIn