Project Description
I need an AI‐powered, ChatGPT-style assistant that plugs directly into my Tally Prime data and gives me conversational answers about every corner of the business: current stock levels, inventory movements, daily sales and revenue, outstanding debtors and creditors, purchase details, even balance-sheet snapshots. The same assistant should also carry out routine actions for me, such as reading a supplier’s PDF, converting it into a Tally-ready sales or purchase invoice, and then emailing statements or invoices to customers right from the system.
I already run everything on Tally Prime, so the project centres on:
• Securely connecting the language model to my live Tally company so it can fetch and post data without breaking compliance.
• Building a chat interface (web or desktop overlay) where I can type or speak questions like “How many units of Item-X are on hand?” or “Show me last quarter’s debtor aging,” and receive clear, formatted answers.
• Automating task flows: PDF-to-invoice creation, scheduled emailing of customer statements, and any similar repetitive jobs we identify during development.
A working proof of concept that handles the queries above and runs these automations end-to-end inside a sandbox company will be the first milestone; we will then refine accuracy, security permissions, and UI polish before moving to production.
Please tell me which tech stack you prefer for the Tally integration and your experience with large-language-model orchestration so we can get started right away.