Treekipedia

DeepTrees is an AI-driven, agent-centric platform that revolutionizes tree research. When a user initiates a research request, our AI research agent (powered by AgentKit) scours the web via the Perplexity API to gather unstructured tree data. GPT‑4o then refines this raw data into a standardized JSON format that fits our database schema. The structured research is stored in PostgreSQL and uploaded to IPFS, generating a unique CID. A custom AgentKit action subsequently triggers an on‑chain EAS attestation on Base L2 to verify the data. Finally, two NFTs are minted—an NFTree (ERC‑721) to the central repository wallet and a Contreebution NFT (ERC‑1155) to reward the contributor. This end‑to‑end process transforms raw, scattered information into actionable, verified knowledge.

DeepTrees automates tree research through a modular, end‑to‑end workflow:

AI-Powered Data Gathering: The AI research agent (powered by AgentKit) initiates a search using the Perplexity API and then employs GPT‑4 to convert raw, unstructured tree data (from scientific and common names) into a standardized JSON format.

Data Storage & Persistence: The structured JSON is stored in a PostgreSQL database and simultaneously uploaded to IPFS via Lighthouse, which returns an immutable CID as a reference.

On‑Chain Verification & NFT Minting: Using AgentKit, our system calls the attestation module to perform an EAS attestation—recording the IPFS CID on Base L2—and then calls our NFT minting module to create two NFTs:

NFTree (ERC‑721): Sent to the Treekipedia wallet. Contreebution NFT (ERC‑1155): Sent to the contributor’s wallet. Unified API Response: The backend (Node.js/Express on a DigitalOcean VM) orchestrates the entire process and returns the final structured JSON—complete with research data, the IPFS CID, and on‑chain transaction details—to the Next.js frontend, where users view the verified research.