How Vorim helped teams build trust into their AI agents.
“Our agents handle real financial and operational data across FinFindr and Kent AI. Vorim gives every one of them an identity and a signed record of what they did, so we can prove it, not just claim it.”
×VorimBinding a verified human to an accountable agent
TruAnon proves the human. Vorim proves the agent. How the integration binds the two and signs both identities into a single record, with quotes from Jesse Tayler and Kwame Nyantakyi.
×VorimReconciling an authorised action with its settlement
Vorim proves the agent was authorised and signs a portable receipt. Zeko binds a privacy preserving commitment to that receipt into the settlement evidence path, so the outcome can be reconciled independently. Verification still never requires a chain.
Read the write-up
FinFindr
FinFindr is an AI-powered opportunity lab. It continuously scans operations, marketing systems, and capital flows across whole industries to find where money is being lost to inefficiency, then designs and deploys agentic solutions to fix them. As the discovery engine of the Kent AI ecosystem, its agents act on real businesses' operational and financial data.
When agents scan and act on a business's operations and capital flows to root out inefficiency, the company on the other side needs to know exactly what each agent was allowed to touch, and trust a record of what it actually did, without taking the platform's word for it.
Scoped identity per discovery agent
Vorim gave each scanning and solution-design agent its own cryptographic identity and least-privilege permissions, so an agent analyzing capital flows could only touch what its task allowed.
Offline-verifiable audit
Vorim signed every agent action at the source, producing a tamper-evident record a client can verify independently, with Vorim nowhere in the trust path.
Runtime control on high-stakes actions
Vorim let consequential actions be gated, modified, or escalated for human approval before they executed, so always-on discovery never meant unsupervised authority over a client's systems.
Kent AI
Kent AI is the operating layer that turns the opportunities FinFindr discovers into fully functional systems. Its first implementation, Kent HOA, takes on the $100B+ HOA and property-management market, a sector that is fragmented and inefficient by default, unifying operations, finance, maintenance, and communication into a single intelligent system. That means agents handling residents' money, records, and maintenance across many communities.
Running real operations and finance for HOAs through agents means being able to answer for every action: a clear permission boundary per agent, and a tamper-evident record an HOA board, resident, or auditor can trust for disputes, incidents, and compliance.
Identity for every operating agent
Vorim gave each agent a first-class cryptographic identity, so every action (a payment, a record change, a maintenance dispatch) was attributable to a specific, verifiable agent.
Fine-grained permissions
Time-bounded, scoped grants enforced before an agent acted, checked in under five milliseconds. An agent could be cleared for one community's finances and nothing else, with no meaningful latency added.
Signed audit trail
A tamper-evident record of every action meant a board dispute or a resident's question was answered with proof, not “trust our logs”.
Runtime control on high-stakes actions
Vorim let consequential actions be gated, modified, or escalated for human approval before they executed, so always-on discovery never meant unsupervised authority over a client's systems.
Listd
Listd is a property services platform where agents, landlords and property managers book inventories, inspections, cleaning and other property work in one place. Verified pros, real-time updates and evidence-backed delivery, with payments held in escrow until the job is done.
When agents list properties and move money on a landlord's behalf, the landlord is trusting software with someone else's home and someone else's deposit. Escrow only releases correctly if the action that triggered it was genuinely authorised, and an evidence-backed job is only as good as the proof that the right party produced the evidence. That needs a permission boundary per agent and a record the landlord, the pro and the platform can each check independently.
Identity for every agent that lists or transacts
Vorim gave each agent acting on a property or a payment its own cryptographic identity, so a listing change, an inspection record or an escrow release is attributable to a specific, verifiable agent rather than to the platform in general.
Scoped permissions before the action
Time-bounded, least-privilege grants checked before an agent acts, in under five milliseconds. An agent cleared to update one property's inventory cannot reach into another landlord's payments.
Signed audit trail on property and money
Every action is signed at source into a tamper-evident record, so a dispute over a deposit, a condition report or a released payment is answered with verifiable evidence rather than with platform logs.