How to Send Microsoft Teams Messages with Python

In both cases, you should end up with a DataLab workbook with your business logic all coded up:

3. Sending a Message to Teams

So, Teams webhook: check. Business logic: check. To integrate the Python notebook with Microsoft Teams, we need to do two things: store the webhook URL in the DataLab workbook and write some Python code to send a message to the Teams channel.

Store the webhook URL

Our Python notebook needs the webhook URL to send the message to the right Teams channel. You could just copy and paste this URL inside a code cell of your notebook. However, this is not a secure approach and is not recommended. The webhook URL is kind of secret. If you share your notebook with someone else, they can see the webhook URL, and they can spam your Teams channel!

This approach would also require you to repeat this copy-pasting in every workbook that you want to send Teams messages with. If you want to update the webhook URL for some reason, you’ll have to go through and edit every workbook.

A more secure and scalable approach is storing the webhook URL token in a so-called environment variable. When you connect this environment variable, it will become available in your Python session. In your workbook:

• Click the “Environment” tab on the left-hand side
• Click on the plus icon next to “Environment Variables”
• In the “Add Environment Variables” modal:
• Set “Name” to MS_TEAMS_WEBHOOK_URL
• In the “Value” field, paste in the Webhook URL that you saved earlier.
• Set the “Environment Variables Set Name” to “Microsoft Teams” (this can be anything, really)

After filling in all fields, click “Create,” “Next,” and finally, “Connect.” Your workbook session will restart, and MS_TEAMS_WEBHOOK_URL will now be available as an environment variable in your workbook. You can verify this by creating a Python cell with the following code and running it:

import os
webhook_url = os.environ["MS_TEAMS_WEBHOOK_URL"]
print(webhook_url)

If you want to reuse the same services account credentials in another workbook, you don’t need to set up the environment variable again: you can reuse the environment variable in other workbooks.

Python code to send the Microsoft Teams message

With the MS_TEAMS_WEBHOOK_URL locked and loaded, we can now write Python code to send a message to Teams! For this, we’ll use the open-source pymsteams package, which provides a convenient wrapper for the pretty intricate API to send messages to Teams. Let’s install it and import it, along with the os package to read out environment variables:

!pip install pymsteams
import pymsteams
import os

Next, we can create a so-called ‘connector card’ – a fancy word for a rich text message in Teams –, add some text to it, and send it:

card = pymsteams.connectorcard(os.environ["MS_TEAMS_WEBHOOK_URL"])
card.text("Hello Teams!")
assert card.send()

Run this code! You should see this message appear in the channel:

Awesome! This was a meaningless sample message, but you can also create a message based on the result of your business logic, as we did in the sample workbook:

card = pymsteams.connectorcard(os.environ["MS_TEAMS_WEBHOOK_URL"])
card.title("Weather Forecast")
card.text("\n".join([
    f"Weather for {CITY} in the next three days:",
    f"- Min temperature: {weather_df['temperature_2m'].min()} {unit_temperature}",
    f"- Max temperature: {weather_df['temperature_2m'].max()} {unit_temperature}", 
    f"- Total precipitation: {weather_df['precipitation'].sum():.2f} {unit_prec}"
]))
assert card.send()

With the following result:

Finally, you can also leverage the DC_WORKSPACE_ID environment variable to include a button that links to your workbook case you want to include an easy way to go from Teams to the DataLab workbook that’s powering this message:

card = pymsteams.connectorcard(os.environ["MS_TEAMS_WEBHOOK_URL"])
card.title("Weather Forecast")
card.text("\n".join([
    f"Weather for {CITY} in the next three days:",
    f"- Min temperature: {weather_df['temperature_2m'].min()} {unit_temperature}",
    f"- Max temperature: {weather_df['temperature_2m'].max()} {unit_temperature}", 
    f"- Total precipitation: {weather_df['precipitation'].sum():.2f} {unit_prec}"
]))
card.addLinkButton("View detailed forecast", f"https://www.datacamp.com/datalab/w/{os.environ.get('DC_WORKSPACE_ID', 'invalid')}/edit")
assert card.send()

With the following result:

Pretty neat, right?!

4. Automate!

You've crafted a Python script that performs a useful analysis and dispatches a summary of this analysis to Teams, including a link back to the complete notebook for those interested in more details. Awesome! Now, imagine if you could set this notebook to execute automatically, daily or weekly, keeping you constantly updated with fresh results. Well, the great news is, you can! DataLab offers the feature of scheduled runs to make this possible.

• In the menu bar at the top, click Run > Schedule Run
• Configure the schedule as you see fit:
• Set up the notebook to run automatically on a daily or weekly basis. For weekly schedules, pick the exact day and time.
• Add DataCamp users that you want to be informed every time the notebook finishes (successfully or with a failure). It’s helpful to do this for failures, so you can dive in and fix when an issue occurs.

If configured like in the screenshot above, the notebook will now run at 9:30 AM local time every day, which means that a message will be posted in the specified Teams channel every morning.

To automate or not to automate

Not every task is a good candidate for automation. It depends on how often you do it, how long it takes to do it manually, and the time it would take to create a Python script to automate it. For instance, consider a yearly report that takes 20 minutes to compile by hand. If it takes you half a day to write a script for it, it'll be many years before you break even on the time spent automating!

Here's a summary in table form:

So if a task that you had to do daily took 5 minutes before, you can spend up to 6 days to try and automate that task. If you spend less time than that automating, you’ll have gained time (counted across five years).

Conclusion

Setting up DataLab to send messages to Microsoft Teams involves navigating through multiple screens initially. But once you've got it all set up with the webhook URL in DataLab, sending a message is just about adding a few lines of Python code – it's really straightforward! The best part? You can automate almost any Python script and directly notify your stakeholders right where they are: on Teams.

Gone are the days of setting reminders for repetitive, mundane tasks. Welcome to an era of efficient, automated workflows where you don't need to switch between tools. All it takes is embracing DataLab.

Remember, If you need a refresher on writing Python code, our Introduction to Python course is the ideal place to start.