Hello everyone,
I have been building something new for a while, but I haven’t shared much about it publicly yet.
It’s called Veyrivo.
Veyrivo is a work marketplace designed around a simple problem: working with someone online still requires a significant amount of trust between people who may have never met each other.
A client needs confidence that the work they are paying for will actually be delivered.
A professional needs confidence that after doing the work and providing what was agreed upon, they will actually get paid.
Traditional freelance platforms try to solve this with profiles, ratings, messaging and centralized escrow. Those things are useful, but I wanted to explore a system where the agreement itself, the milestones, the evidence of work completed and the payment lifecycle are much more structured.
That is what I am building with Veyrivo.
Veyrivo allows clients to discover professionals and allows professionals to find work opportunities, but the product goes beyond simply matching both sides.
A project can be broken down into clearly defined milestones with acceptance criteria and expected evidence before work begins.
The general flow looks like this:
Discover → Discuss → Proposal → Milestones → Agreement → Fund → Work → Submit Proof → Review → Release Payment
Instead of relying only on messages and informal promises, Veyrivo is being designed so that the important parts of a work relationship are represented as structured actions.
Clients can define what they expect.
Professionals know what they need to deliver.
Proof can be attached to milestone submissions.
Both sides can follow the state of an agreement.
And payments can be tied directly to the agreed workflow.
But one of the most important things about Veyrivo is actually what sits underneath it.
Veyrivo is being built on PactAgent
Some members of the Nervos community may already know PactAgent.
PactAgent was my submission to the Claw & Order: CKB AI Agent Hackathon, where it won 1st Place.
The original PactAgent was built as a milestone-based escrow and payment application on CKB.
A client could create an agreement, divide work into milestones and fund the agreement. A worker could complete the work and submit proof, while an autonomous agent monitored the agreement lifecycle, checked deadlines, reviewed submitted evidence, progressed milestones and assisted with outcomes such as payment, refund or disputes.
AI was intentionally kept advisory rather than being given unrestricted control over funds.
That hackathon version proved the concept.
But after the hackathon, I started asking a different question:
What if PactAgent should not just be one application?
What if other applications could use the same agreement, milestone, proof, escrow and settlement infrastructure?
That question led to a major change in the direction of the project.
PactAgent is now infrastructure
Over the past development cycles, I have been refactoring PactAgent from the standalone product that won the hackathon into reusable infrastructure that other applications can integrate with.
The new PactAgent architecture is app-scoped and exposes its capabilities through an API.
It now provides infrastructure around areas such as:
-
Apps and projects
-
Scoped API keys
-
Agreements
-
Milestones
-
Escrow lifecycle management
-
Proof submission
-
Reviews
-
Disputes
-
Transactions and settlement
-
Events
-
Webhooks
-
Audit logs
-
Idempotent operations
-
Sandbox/testing workflows
-
Operational and administrative tooling
This means an application no longer needs to become PactAgent in order to benefit from PactAgent.
It can build its own product, its own users, its own business rules and its own user experience while using PactAgent underneath for the trust and agreement infrastructure.
And Veyrivo is the first product being built this way.
How Veyrivo uses PactAgent
Veyrivo owns the marketplace experience.
Things such as professional profiles, portfolios, discovering talent, discovering jobs, proposals, messaging, reputation and the overall hiring experience belong to Veyrivo.
Once two people decide to work together, however, PactAgent becomes an important part of the underlying workflow.
Veyrivo can translate the agreed work into structured PactAgent agreements and milestones.
When a milestone is funded, PactAgent manages the corresponding escrow lifecycle.
When a professional completes a milestone and submits evidence, that proof can move through PactAgent’s proof and review infrastructure.
PactAgent can assist with evaluating whether the required information has been provided while still maintaining human checkpoints where important decisions—particularly decisions involving funds—need confirmation.
The infrastructure can then track the agreement toward approval, payment, refund or dispute depending on what happens during the work.
Events generated throughout that lifecycle can be sent back to Veyrivo through webhooks so the user sees a simple marketplace experience without needing to understand everything happening underneath.
From the user’s perspective, they are simply using Veyrivo.
From the infrastructure perspective, PactAgent is coordinating much of the trust lifecycle behind the transaction.
That separation is intentional.
Why build Veyrivo first?
Building infrastructure without a real application consuming it can make it easy to design APIs around assumptions.
Veyrivo gives me a real product against which PactAgent can be tested.
If creating an agreement is unnecessarily difficult, Veyrivo exposes it.
If the milestone lifecycle is missing an important state, Veyrivo exposes it.
If proof submission is awkward, Veyrivo exposes it.
If webhooks, idempotency, escrow operations or dispute handling don’t behave correctly in a real application workflow, Veyrivo exposes those problems as well.
So I see Veyrivo not only as a product, but also as PactAgent’s first serious reference implementation.
The goal is for PactAgent to eventually make these capabilities available to other builders without requiring everyone to rebuild the difficult parts of milestone agreements, escrow, proof verification and settlement themselves.
Where Veyrivo is today
Veyrivo is still under active development and has not been publicly hosted yet.
There is still work to do before I consider it ready for people to start using.
But I also don’t want to wait until everything is finished before talking about what I am building.
One of the things I want to do differently this time is share more of the development process with the community—the architecture decisions, what works, what doesn’t, how PactAgent evolves while being used by a real application, and eventually what happens when Veyrivo starts onboarding real users.
PactAgent started as a hackathon project.
It won the Claw & Order: CKB AI Agent Hackathon.
Now it is becoming infrastructure.
And Veyrivo is the first application that will show what can actually be built on top of it.
This is still early, but I think this next stage of the project is much bigger than simply continuing the original PactAgent application.
More updates on both Veyrivo and the PactAgent infrastructure will follow as development progresses.
Veyrivo — Work with trust. Deliver with proof.
https://github.com/Ajayfrizzy/veyrivo.git
