Business Result

<aside> 💡

Problem: In a small business, customer information lives in the owner's head, phone notes or chat history. A full CRM is too expensive and too complex at this scale.

Before: Saving a customer meant opening a spreadsheet, splitting the full name into columns by hand and typing a note. Nobody did it on the go.

What we built: A minimal CRM on Make: the owner sends the bot a description of a customer in plain language, AI splits it into last name, first name, patronymic and a note, and the finished record with a timestamp appears in Google Sheets.

Result: Adding a customer to the database now takes one Telegram message. Records are structured and timestamped, and the system needs neither code nor a paid CRM.

</aside>

A no-code Make automation: a Telegram bot receives a free-text customer description, OpenAI structures it into fields, and Google Sheets works as a simple customer database.

Project Details

Context:

• Business/industry: Small business / made-to-order bakery
• Current process: Customer details are recorded chaotically or not at all, with no single database.

Problem:

• No single place to store customers and notes about them.
• Filling in a spreadsheet by hand is inconvenient, so it keeps getting postponed.
• A full CRM is overkill in features and price at this scale.

Solution

<aside> 🧠

A Telegram bot receives a message like "Максименко Ігор Валерійович, регулярно замовляє торти" (orders cakes regularly). Following its instructions, the OpenAI module splits the text into last name, first name, patronymic and note and returns them as JSON. The JSON module turns the response into fields, and Google Sheets adds a row with a creation timestamp. There is nothing technically complex to configure: every module comes ready-made in Make, and data flows between them through field mapping, with no code.

Tools used: Make, Telegram Bot API, OpenAI API, Google Sheets.

</aside>

Solution logic:

1. Telegram Bot → Watch Updates receives a message with the customer description.