QuickLedgerBooks

QuickLedgerBooks is an AI-powered bookkeeping and crypto payments platform for small businesses. Instead of manually managing vendors, invoices, payments, receipts, and transaction records, users can simply describe what they want in natural language.

The platform combines AI agents, ENS identities, Ledger-secured approvals, and automated bookkeeping workflows into a single experience. Users can save merchants and customers, create invoices, prepare payments, generate receipts, and maintain an audit trail through conversational prompts.

When a user requests a payment, the AI agent first identifies the merchant, retrieves payment history, resolves ENS names into wallet addresses, performs risk checks, and prepares a payment review summary. The system never sends funds automatically. Every outgoing transaction requires explicit approval through a Ledger Nano S Plus hardware wallet, ensuring that AI can assist with payments without ever controlling funds.

After Ledger approval, the payment is broadcast to Ethereum Sepolia. The system automatically records the transaction, generates a PDF receipt, emails the receipt to the merchant, and stores a permanent bookkeeping record in MongoDB. This creates a complete end-to-end workflow from AI instruction to payment settlement and accounting documentation.

QuickLedgerBooks demonstrates how AI agents can safely interact with financial workflows when paired with hardware-wallet security and human approval.

QuickLedgerBooks is built using Next.js, TypeScript, MongoDB, LangChain, Google Gemini, ENS, Ethereum Sepolia, Ledger hardware wallets, and Privy authentication.

The frontend is built with Next.js and Tailwind CSS, providing dashboards for merchants, customers, payments, receipts, and AI agent interactions. Privy is used for wallet authentication and onboarding.

The AI layer is implemented using LangChain tools and Google's Gemma/Gemini models. Custom tools allow the agent to create merchants, manage customers, generate invoices, prepare payments, retrieve payment history, resolve ENS names, and update bookkeeping records. Every tool invocation is logged and stored to provide transparency into agent actions.

MongoDB stores merchants, customers, payments, invoices, receipts, and agent activity logs. This provides a complete audit trail of every action performed by the system.

For payments, merchants can be identified by ENS names instead of manually entered wallet addresses. The platform resolves ENS names using Ethereum mainnet RPC endpoints and uses the resolved wallet address when preparing payment transactions.

Security is provided through Ledger Nano S Plus integration. The AI agent can prepare transactions, but cannot move funds. Users must physically review and approve transactions on the Ledger device before any blockchain transaction is broadcast.

After a payment is confirmed, the platform automatically generates a PDF receipt using pdf-lib, stores the receipt on the application server, sends it by email through Nodemailer and Gmail SMTP, and saves the receipt metadata in MongoDB. This creates an automated bookkeeping workflow that combines AI automation with hardware-wallet security and verifiable blockchain settlement.