HisabMate: A Local-First, AI-Assisted Digital Bahi-Khata for Indian Microbusinesses

Description and use

HisabMate is a local‑first digital ledger for Indian microbusinesses that uses a local LLM to propose transactions, ensures exact integer‑paise calculations, offers CSV export, and keeps data on the owner’s machine.

Problem

Traditional handwritten credit ledgers in Indian microbusinesses are fragile, prone to smudging, missing entries, and disputes over balances.

Outcome

HisabMate provides a reliable, searchable, and exportable digital ledger for microbusinesses while keeping data local and avoiding third‑party cloud dependencies.

Workflow

• Install frontend dependencies with npm
• Run backend to auto‑create SQLite schema
• Configure Ollama with llama3.2:3b
• Use the web UI to add customers and record transactions
• AI proposes transaction details from natural language input
• Confirm or reject AI proposals before committing to ledger
• Export filtered ledger to CSV
• Review ledger via customer‑specific or daily views

Evidence

HisabMate is a local‑first, AI‑assisted bahi‑khata for Indian microbusinesses.

Source

The model proposes and the code decides.

Source

Integer paise, never floats.

Source

Model output is a proposal only.

Source

I used GitHub Copilot Chat in VS Code, primarily in Agent mode, as a coding assistant throughout HisabMate's development.

Source