Skip to main content
How to train the bot on tabular data with CSVs

Learn how you can train the Threado AI bot on tabular data using CSVs and ask queries on Slack.

Pramod Rao avatar
Written by Pramod Rao
Updated over 10 months ago

It can be difficult and not to mention time-consuming to search through across CSVs. With Threado AI, you can train the bot on tabular data and ask relevant queries from the CSV directly.

Train the bot on CSVs

You can also train your bot on tabular data by uploading CSVs.

πŸ“Œ Note: Upto 15 files can be uploaded for training.

After the bot has been trained, you can ask questions based on the tabular data on the Slack channels where the bot has been installed.

πŸ“Œ Note: Since the bot only answers queries on Slack channels, you will have to install the bot on Slack if you haven't already. To learn more about this, read this article.

The bot will answer the queries based on the corresponding data it finds in the CSV.

What are some ways in which you can leverage tabular data through the bot?

Get customer information directly from the bot

If you have all your customers' data in an Airtable or a Google sheet, it can be exhaustive to have to search through individual customers every time you need to reference information.

Train the bot on a CSV of this data and ask questions directly to the bot. Example - if you want to know which customers are on the 'Starter' plan, query this on the Slack channel and the bot will give you all the information.

Keep employee information handy at all times

Train the bot on a sheet with all your organization's employee information. This can be particularly helpful for HRs or anyone looking for any information on an employee i.e. email, which department they belong to, etc.

Have better coordination with your customer support team

If your team maintains a timetable in a tabular data format, it can be easy to coordinate and be aware of your team's whereabouts. An example could be - querying the bot to find out who is working the evening shift on a particular day.

Did this answer your question?