✍️ Key Tips for an Effective Prompt
Developing an effective prompt can make the difference between obtaining precise and clear results, or facing confusing and unhelpful responses. Make sure your prompt is well-structured and optimized by following these recommendations:
Information Quality: The foundation of any effective prompt is the quality of the data provided. Ensure that it is well-written, free of spelling errors and structured clearly (use headings and lists to organize information).
Organization: Divide the information into clear and concise sections that delimit the data to train. This will help the bot process information more effectively.
Appropriate Format: Use file formats that are compatible with the platform (such as plain text in PDF format or the one specifically required). Avoid using boxes or other visual elements that may hinder the bot's data reading.
Relevant Q&A: Consider the questions that are frequent for your customers and those that should be answered in a certain way. Train the bot by including these questions and answers in a dedicated section to reduce incoherent responses.
Image Handling: If you need to include images, upload them to the cloud and generate a link with the appropriate ending (.png, .jpg). Then include these links in the Q&A section. General documents are uploaded from your computer in the 'files' section and should not be linked from the internet.
🏗️ Detailed Prompt Structure
To provide the bot with clear context about how it should act, we recommend dividing your prompt into the following fundamental sections:
Context: Describe how the bot should behave, its tone of voice and what it should report. (Example: "Act as an expert in car sales for company X, and respond to customer inquiries in a friendly manner. If someone asks information unrelated to the company, respond that you don't have access to that information").
Format: Describe how the bot should respond in terms of the visual presentation of its responses. (Example: "Respond in a message of no more than 3000 words, without emojis, and organize the information into lists and paragraphs of no more than 3 lines").
Restrictions: Highlight everything that you do not want the bot to respond to or do. (Example: "If someone asks about competitor products, respond that you cannot provide that information. You cannot be creative with responses").
🧠 Validation of Consistency in Knowledge Bases
When a bot uses a structured knowledge base to respond to queries (such as prices, descriptions or product availability), it is essential to ensure that all fields are correctly updated. A small misalignment can cause incorrect responses in production.
✅ Checklist for Publishing Updates
Before saving or publishing changes to your base, make sure to review the following:
Did you update all relevant columns? (📝 Ex.: name, description, price, stock, URL, etc.)
Were the data modified in the correct version of the month or bot?
Was a review of the complete dataset (rows and columns) done before saving?
Did you test the behavior in the bot by simulating the intent or keyword?
⚠️ Common Errors You Can Avoid
Frequent Error | Why Does It Occur? | How to Avoid It |
Only the Product Description is Updated | A cell in the row is edited without checking other related columns. | Use filters to view the complete content by product before saving. |
The Correct Document is Edited, But Not Published | The knowledge base remains as a draft or is not linked to the correct bot. | Confirm in the bot menu that the active Knowledge Base is the appropriate one. |
Duplication Without Cleanup | A dataset is duplicated without deleting previous values. | Review and clean the columns before duplicating or copying the information. |
🧪 Real Example:
A customer reported that the price of "item 1" was incorrect. The description had been updated, but the price column maintained the previous value.
👉 Solution: The price field was updated in the base corresponding to the month and verified in the flow. As a result, the bot began to respond correctly with the updated price.
🧩 Final Recommendation
If you work with multiple monthly versions or campaign datasets (databases), implement a clear naming convention and an internal control table that indicates:
Date of last edit.
Published version.
Responsible user.
Which specific fields were modified.
Train your AI like your best agent! 🚀
Structuring detailed prompts and keeping your knowledge bases free of errors ensures that your bot provides safe, precise and highly reliable responses to your customers. ✅