ARCA

AI native outsourcing marketplace that matches work with verified humans, uses AI to break down and evaluate tasks, and creates work provenance so companies can trust the work before they pay.

Comment 3

Inspiration

Outsourcing is built around a simple assumption: if a company pays someone to do work, it should be able to trust that the work was actually completed and that the result is genuine.

That assumption breaks down when work is increasingly produced with AI. A freelancer can submit an output, but the client often has little visibility into how the work was produced, whether the claimed process actually happened, or whether the result meets the requirements.

We built ARCA around one principle:

Trust the worker → Verify the work.

ARCA is an AI-native outsourcing marketplace designed to make outsourced work more verifiable.

What ARCA does

A company starts by describing a project in plain language.

ARCA decomposes that request into structured tasks and connects the work to Integration Contracts. Workers can then compete for the available tasks and submit their completed work.

The important difference is what happens after submission.

ARCA verifies the work before settlement. The platform records AI actions and work provenance, performs verification checks, and uses the resulting evaluation to determine how rewards are distributed.

The goal is not simply to create another freelancer marketplace. It is to create an outsourcing workflow where verification is part of the product itself.

How we built it

ARCA is built as a full-stack application with a React Native/Expo mobile experience and a Fastify-based API.

The project includes:

  • AI-assisted task decomposition
  • Worker and company workflows
  • Work verification and evaluation
  • AI provenance and action records
  • Sui-based escrow/settlement architecture
  • World identity integration
  • MultiBaas payout integration
  • ARCA Pro subscription infrastructure using RevenueCat
  • A responsive web experience for demonstrating the complete workflow

For the hackathon environment, external blockchain, identity, AI, and settlement interactions are represented through deterministic mock/demo integrations where live credentials or production infrastructure are not available. The verification, evaluation, ranking, reward calculation, and hashing logic are implemented in the project rather than being presented as simulated results.

RevenueCat is integrated into the ARCA Pro subscription flow using the RevenueCat Test Store for development and demonstration.

What we learned

The biggest lesson was that building an AI-native marketplace is not only an AI problem.

The difficult part is connecting multiple systems into one coherent workflow: task decomposition, worker execution, verification, provenance, identity, settlement, and monetization.

We also learned that trust needs to be designed into the workflow, rather than added as a feature after the marketplace is built.

Challenges

One of our biggest challenges was making the different parts of the system work together while keeping the demo understandable.

We had to design a workflow that could demonstrate the complete lifecycle of outsourced work without depending entirely on production credentials or live external infrastructure.

We also spent significant effort making the application consistent across its mobile, web, API, and documentation layers.

What's next

ARCA is currently a prototype, but the architecture is designed around a larger marketplace.

The next stage would be connecting production AI providers, identity verification, blockchain settlement, and real users, followed by expanding the verification system and introducing production subscription and marketplace monetization.

Our long-term goal is simple:

Make outsourced work verifiable by default.

Built With

Share this project:

Updates

Submission history