diff --git a/assets/manage/data/data_management.png b/assets/manage/data/data_management.png
new file mode 100644
index 0000000000..a1d96d9f5a
Binary files /dev/null and b/assets/manage/data/data_management.png differ
diff --git a/assets/program/on-robot.png b/assets/program/on-robot.png
new file mode 100644
index 0000000000..8e619022dc
Binary files /dev/null and b/assets/program/on-robot.png differ
diff --git a/assets/program/remotely.png b/assets/program/remotely.png
new file mode 100644
index 0000000000..f4fa647513
Binary files /dev/null and b/assets/program/remotely.png differ
diff --git a/docs/manage/data/_index.md b/docs/manage/data/_index.md
index bc611aa828..a3d3beb0b5 100644
--- a/docs/manage/data/_index.md
+++ b/docs/manage/data/_index.md
@@ -11,7 +11,9 @@ aliases:
 # SME: Aaron Casas
 ---
 
-Data Management is built into Viam allowing you to capture, sync, [view](view/), [label](label/), and [export](export/) data from your robots, as well as use Machine Learning features to enhance your robot's capabilities.
+Data Management is built into Viam allowing you to capture, sync, [view](view/), [label](label/), and [export](export/) data from your robots, as well as use [Machine Learning](/manage/ml/) features to enhance your robot's capabilities.
+
+![Data is captured on the robot, uploaded to the cloud, and then deleted off local storage.](/manage/data/data_management.png)
 
 {{< cards >}}
 {{% manualcard %}}
@@ -26,13 +28,12 @@ Data Management is built into Viam allowing you to capture, sync, [view](view/),
 Once you have configured data capture and cloud sync, you can <a href="/manage/data/view/">view</a> and <a href="/manage/data/export/">export</a> your data.
 {{% /manualcard %}}
 
-{{% manualcard link="../ml/" %}}
-
-<h4>Machine Learning</h4>
-<p>Use collected data to train machine learning models or use existing models on your robots.</p>
-{{% /manualcard %}}
 {{< /cards >}}
 
 ## Next steps
 
-For a comprehensive tutorial on data capture, synchronization, and accessing captures data, see [Intro to Data Management](../../tutorials/services/data-management-tutorial/).
+For a comprehensive tutorial on data capture, synchronization, and accessing captures data, see the following tutorial:
+
+{{< cards >}}
+{{% card link="/tutorials/services/data-management-tutorial/" %}}
+{{< /cards >}}
diff --git a/docs/program/run.md b/docs/program/run.md
index 5cefd29f30..69c96aa5dd 100644
--- a/docs/program/run.md
+++ b/docs/program/run.md
@@ -27,6 +27,9 @@ You can find it at the top of the robot's **Control** tab.
 ## Run Code Remotely
 
 Most of the time, as long as both you and your robot are connected to the internet, you will want to run code to control your robot remotely.
+
+{{<imgproc src="/program/remotely.png" resize="800x" declaredimensions=true alt="A client connecting remotely to a robot">}}
+
 The advantage of this method is that your robot and your computer do not have to be connected to the same WAN/LAN to issue control commands.
 You can remotely control your robot with any application you implement from anywhere in the world.
 For example, you can run code on your personal computer, creating a client [session](/program/apis/sessions/), where the code running on that computer sends instructions to your robot's `viam-server` instance over the internet.
@@ -73,6 +76,8 @@ This is useful because as long as that computer is able to establish a network c
 
 In case you run [PID control loops](https://en.wikipedia.org/wiki/PID_controller) or your robots have intermittent network connectivity, you can ensure this does not interfere with the code's execution, by running the control code on the same board that is running `viam-server`.
 
+{{<imgproc src="/program/on-robot.png" resize="800x" declaredimensions=true alt="A client running on a robot">}}
+
 When connecting to a robot using the connection code from the [code sample tab](/program/#hello-world-the-code-sample-tab), a [client session](/program/apis/sessions/) automatically uses the [most efficient route](/program/connectivity/) to connect to your robot, which means the favored route for commands will be over localhost.
 
 ## Run code automatically