Summary (3–4 sentences):
Onchain Quest is a prototype P2E service that converts short daily quizzes into on-chain micro-rewards. Users complete a one-minute off-chain quiz, obtain an EIP-712 signed claim ticket, and submit that ticket via a relayer (meta-transaction) to receive ERC-20 tokens or occasional ERC-721 NFTs — all with minimal gas friction. AI assists with personalized quiz generation and fraud detection, while blockchain provides transparent, auditable reward settlement.
Clearly describe the real-world or technical problem you are addressing.
Many reward or learning apps want to convert lightweight user actions (micro-tasks) into verified on-chain events, but onboarding friction (wallet setup, gas fees) and UX complexity keep non-crypto users away. Even when users complete off-chain tasks, they often skip on-chain claims because of gas cost or confusion.
What are the limitations or inefficiencies of existing solutions?
Existing P2E or micro-reward systems either force users to hold native tokens (gas) or rely on custodial off-chain accounting that lacks transparency. Solutions that subsidize gas are costly and hard to scale without careful economics. Many projects also lack automated anti-abuse mechanisms tailored to low-value, high-frequency tasks.
Why is this problem worth solving?
Reducing onboarding friction and reliably converting lightweight engagement into on-chain activity unlocks broader adoption of blockchain UX, enables transparent incentive systems for education / micro-learning / community tasks, and provides verifiable proof-of-participation for DAOs, research, or loyalty programs.
What is your proposed solution, and how does it work?
Onchain Quest combines a minimal UX with a meta-transaction flow: users complete a short off-chain quiz → the backend issues an EIP-712 claim ticket → the user signs the ticket (or receives a server-issued signed ticket) → the ticket is submitted to a relayer which pays gas and calls claimReward(...) on the smart contract. The contract verifies the signature and claimId to prevent replay and transfers the reward.
How are AI & Blockchain combined in your approach?
AI for content & personalization: a light AI module generates daily quiz items, tailors difficulty per user, and rotates content to maintain engagement.
AI for fraud detection: a lightweight model scores suspicious activity patterns (fast repeated completes, identical answers, abnormal timing) so the relayer or backend can flag or throttle claims.
Blockchain for settlement & audit: final reward distribution is settled on-chain (ERC-20 / ERC-721), which provides immutable proof of reward allocation and prevents disputes.
Blockchain provides transparent, tamper-proof settlement of rewards. This is important for verifiability (proof of participation), for enabling trustless distribution of scarce NFT rewards, and for future interoperability with DAOs or on-chain reputation systems.
What problem does blockchain specifically solve in your context?
It removes ambiguity about who received rewards, allows external parties to validate participation, and enables composability (e.g., turning earned tokens into governance weight or NFT access).
If you designed a token, what is its purpose (e.g., incentive, governance, reputation)?
Optional token design: An ERC-20 utility token used for small rewards and a limited ERC-721 series for special achievements. Tokens serve as immediate incentives; future work may add token staking for governance or premium features. (In the current prototype we focus on reward distribution; governance is out of scope.)
Removing gas friction meaningfully increases claim conversion in simulation. The meta-tx design is promising for non-crypto users, but operational and economic constraints (relayer cost, fraud mitigation) are central to real deployability.
What were the biggest challenges?
Designing a secure but low-friction signature flow (EIP-712 nuance).
Balancing believable simulation assumptions against real user variability.
Planning sustainable relayer economics without enabling abuse.
What would you improve or build next if you had more time?
Build an actual Sepolia pilot with a small real user group to validate assumptions.
Implement a production relayer with budget controls and throttling.
Integrate a lightweight ML model (server side) for live bot detection and AB testing for reward sizing.
Conduct a security audit and refine token economics (if tokens are to be persistent).
