dot402

Inspiration

Traditional AI systems are black boxes — they think internally, produce an output, and you never know how the reasoning happened or how much each step cost. At the same time, blockchain payments usually happen at the end of a workflow, not during it.

We wanted to combine these two worlds: a system where every step of the work is visible, accountable, and paid for in real time.

How we built it

We created a backend pipeline that processes tasks in sequential reasoning steps (understand → extract → summarize → finalize).

After each step, the backend triggers a real blockchain payment using the Polkadot API.

Payments use AssetHub so every step appears as a balances.Transfer on-chain.

The frontend visualizes each step along with its payment transaction hash.

If the chain connection fails, the system falls back to demo mode but keeps the reasoning flow intact.

This makes the project simple, reliable, and fully tied to the chain we are working with.

What it does

dot402 takes any user-submitted task, breaks it into multiple reasoning steps, processes them one by one, and after every step it sends a real micropayment on AssetHub using balances.transfer.

This creates:

Transparent step-by-step execution

On-chain proof of work for each step

A simple “pay as you go” flow

A live demo of how autonomous agents could pay for their own operations

How we built it

We created a backend pipeline that processes tasks in sequential reasoning steps (understand → extract → summarize → finalize).

After each step, the backend triggers a real blockchain payment using the Polkadot API.

Payments use AssetHub so every step appears as a balances.Transfer on-chain.

The frontend visualizes each step along with its payment transaction hash.

If the chain connection fails, the system falls back to demo mode but keeps the reasoning flow intact.

This makes the project simple, reliable, and fully tied to the chain we are working with.

Challenges we ran into

Finding the correct RPC for the specific chain (Passet Hub / IBP1) without running into timeouts.

Understanding SS58 prefixes and why using the wrong address format causes runtime panics.

Managing payments safely while still making the demo smooth enough for a hackathon.

Ensuring the backend never crashes even if a payment fails — instead it switches to safe demo mode.

Accomplishments that we're proud of

We built a working proof-of-concept where every reasoning step creates a real on-chain transfer.

The entire workflow is transparent — inputs, reasoning, payments, and results.

First time implementing step-based blockchain payments tied to AI processing.

A clean UI that clearly shows each step and the corresponding transaction.

Fully functional backend pipeline with fallback modes to avoid downtime during demos.

What we learned

How to interact with AssetHub using Polkadot’s API.

How address prefixes work and why they matter.

How to design a system that is resilient to RPC failures.

How to combine sequential reasoning with blockchain payments cleanly.

That even simple micropayment flows need careful chain selection and correct key formats.

What's next for dot402

Adding support for multiple parachains and cross-chain workflows.

Replacing simulated reasoning outputs with real AI models.

Introducing dynamic pricing per step based on complexity.

Allowing third-party agents to plug in their own workflows and earn payments per action.

Adding dashboards to analyze cost per task and efficiency per step.

Challenges we ran into

Accomplishments that we're proud of

What we learned

What's next for dot402

Built With

• next