Building this project presented several steep challenges. The most difficult part was understanding Aztec and its native language, Noir. Coming from traditional programming environments, Noir’s syntax felt minimal and unfamiliar. Even writing basic logical operations or structuring conditionals required a completely different way of thinking all within the context of zero-knowledge circuits. Understanding how Aztec handles public versus private variables, commitments, and storage was intellectually demanding. Another major hurdle was zk identity verification. Integrating zkPassport in a way that allowed private, Sybil-resistant voting without compromising identity was tricky. It took time to grasp how to prove someone was a unique human without actually knowing who they were. Finally, integrating a TEE-based AI agent into a ZK-powered on-chain flow was complex. Making this agent fetch, process, and verify disaster and NGO data and then trigger smart contract calls required careful separation of trusted and untrusted components. The multi-chain bridging experience also posed UX challenges. I overcame these hurdles by diving deep into Aztec documentation, breaking problems into modular steps, seeking support from the community, and iterating constantly until each piece worked.