Skip to main content
Each example includes the brief, target group, and a representative conversation. The conversations are illustrative — actual conversations are generated dynamically per user from the same brief.

​
RLHF: DeFi response ranking

Advertiser: AI lab building a DeFi assistant
Group: Crypto Traders HQ · 4,200 members · DeFi / trading Brief:
We are training a DeFi assistant model and need preference labels from active traders. Ask users to compare two AI responses to a swap-related question and explain which is more helpful and why. We need the reasoning, not just the selection.
Conversation:
Canvas: Before you join — we’re collecting feedback for a DeFi AI model. Which of these responses to “how do I swap ETH to USDC safely” is more helpful, and why? A: “Use a DEX like Uniswap. Connect your wallet, enter the swap amount, and confirm.”
B: “Set slippage to 0.5% or lower. Check the pool’s TVL first — low liquidity means more price impact. Approve the token, then confirm the swap.” User: B, obviously. A doesn’t tell you anything useful Canvas: Agreed — why does the slippage part matter more than people think? User: Because on a big swap on a thin pool you can get wrecked. I’ve seen people lose 5–10% just from not checking liquidity before Canvas: That’s exactly the kind of context we’re looking for. You’re in.
What the advertiser receives: A labeled preference pair with a reasoning explanation from a verified DeFi trader. The second turn was triggered because the first response was short. The expanded answer is the training signal.

​
RLHF: Solidity code review

Advertiser: AI lab building a coding assistant
Group: Web3 Developers · 2,100 members · Solidity / smart contracts Brief:
We need Solidity developers to evaluate AI-generated code snippets for correctness and security. Ask users to spot any issues in a short function. We want to know if they catch the reentrancy risk.
Conversation:
Canvas: Quick question before you join — we’re training a Solidity code assistant. Can you spot any issue with this function? 
function withdraw(uint amount) external {
    require(balances[msg.sender] >= amount);
    (bool success,) = msg.sender.call{value: amount}("");
    require(success);
    balances[msg.sender] -= amount;
}
User: Reentrancy. The balance update happens after the external call Canvas: Correct. How would you fix it? User: Move the balance update before the call. Or use a reentrancy guard. Both work but updating state first is the cleaner pattern Canvas: Good answer. You’re in.
What the advertiser receives: A verified Solidity developer who identified a reentrancy vulnerability and explained the fix. The second turn confirmed depth of knowledge beyond pattern recognition.

​
Agent acquisition: Lending protocol

Advertiser: Aave
Group: Base DeFi Traders · 2,800 members · lending / yield Brief:
We want to reach Base-native users who actively manage yield positions. Ask them where they currently put idle capital and what factors matter when choosing a protocol. After verification, offer a link to our automated lending agent.
Conversation:
Canvas: Quick question — if you had idle USDC right now, where would you put it and why? User: Probably Moonwell or Aave on Base, decent rates and I trust the contracts Canvas: Do you actively manage your positions or set and forget? User: Bit of both. I check weekly, move stuff around if rates shift a lot Canvas: You’re in. Aave has an agent that rebalances lending positions automatically if you want to check it out — link’s in the welcome message.
What the advertiser receives: A verified Base DeFi user with stated protocol preferences and described management behavior.

​
Agent acquisition: Perpetuals trading

Advertiser: A perpetuals trading protocol
Group: Crypto Traders HQ · 4,200 members · DeFi / trading Brief:
We want to reach active perps traders. Ask how they currently manage risk on leveraged positions. No correct answer required — we want to understand how experienced traders think about this.
Conversation:
Canvas: Quick question for a trading protocol we’re partnering with. How do you manage risk when you’re in a leveraged position? User: Stop losses and I never go above 5x Canvas: Do you set stops based on price levels or percentage from entry? User: Usually percentage, around 3–5% depending on the trade. Sometimes key levels if there’s an obvious support Canvas: That’s useful context. You’re in.
What the advertiser receives: A verified active trader with specific risk management behavior described across two turns.

​
Brand research: NFT marketplace

Advertiser: An NFT marketplace
Group: NFT Collectors · 3,100 members · art / NFTs Brief:
We’re building a discovery product for NFT drops and want to understand how active collectors find out about new projects. Ask them what triggers a purchase — not where they browse generally, but what makes them actually buy.
Conversation:
Canvas: Before you join — we’re doing research for an NFT discovery tool. What usually tips you over into actually buying a drop, not just looking? User: Honestly usually it’s someone I trust posting about it. Or if the artist is someone I’ve followed for a while Canvas: More the person behind it than the art itself? User: Both, but if I don’t know the artist the trust has to come from somewhere else. Like if three people I respect are all minting it that’s usually enough Canvas: That’s useful. You’re in.
What the advertiser receives: A first-person purchase decision framework from a verified NFT collector, with social proof behavior described across two turns.

​
Consumer research: Wallet onboarding

Advertiser: A crypto wallet app
Group: Crypto Beginners · 1,800 members · onboarding / education Brief:
We want to understand what friction points new crypto users hit in their first month. Ask what was most confusing when they were getting started. We’re more interested in emotional experience than technical depth — what made them feel stuck.
Conversation:
Canvas: What was the most confusing part of getting into crypto when you first started? User: Honestly gas fees. I didn’t understand why the fee was sometimes more than what I was sending Canvas: Did you figure it out eventually or did you just avoid small transactions? User: Avoided them for a while. Then someone explained it was like a toll road and it clicked Canvas: That’s a really common one. You’re in.
What the advertiser receives: A first-person account of a specific onboarding friction point, plus a behavioral response (avoidance), from a verified beginner-segment user.