diff --git a/.env.example b/.env.example
index 6817eb98d..6c6a4fd03 100644
--- a/.env.example
+++ b/.env.example
@@ -60,3 +60,6 @@ GOOGLE_CLIENT_SECRET=your_google_client_secret
 
 KEYCLOAK_REALM=quepid
 KEYCLOAK_SITE=http://keycloak:9080
+
+# Whether or not to do analytics for interesting events in Case Scores
+QUEPID_PROPHET_ANALYTICS=true
diff --git a/app/views/home/_case_summary.html.erb b/app/views/home/_case_summary.html.erb
index d978a2470..d2955ae20 100644
--- a/app/views/home/_case_summary.html.erb
+++ b/app/views/home/_case_summary.html.erb
@@ -10,7 +10,6 @@
       </p>
       <!-- Look at https://github.com/ankane/prophet-ruby -->
       <%
-      #df = Rover.read_csv("example.csv")    
       data = kase.scores.collect{ |score| {ds: score.created_at.to_date.to_fs(:db), y: score.score, datetime: score.created_at.to_date } }.uniq
       # warning! blunt filter below!
       data = data.uniq { |h| h[:ds] }
@@ -18,7 +17,7 @@
       
       do_changepoints = data.length >= 3 ? true : false # need at least 3...
       
-      if do_changepoints
+      if Rails.application.config.quepid_prophet_analytics && do_changepoints
         df = Rover::DataFrame.new(data)
         m = Prophet.new()
         m.fit(df)
diff --git a/config/initializers/customize_quepid.rb b/config/initializers/customize_quepid.rb
index ac2c2164e..8338b9553 100644
--- a/config/initializers/customize_quepid.rb
+++ b/config/initializers/customize_quepid.rb
@@ -94,3 +94,8 @@
 
 # == Disable redirecting to match the TLS setting
 Rails.application.config.redirect_to_match_search_engine_tls = ENV.fetch('REDIRECT_TO_MATCH_SEARCH_ENGINE_TLS', true)
+
+# == Prophet Analytics Control
+# Prophet tells you interesting things about time series curves and is used on the homepage.
+# It may consume too much memory for your environment, and you may need to disable it.
+Rails.application.config.quepid_prophet_analytics = bool.deserialize(ENV.fetch('QUEPID_PROPHET_ANALYTICS', true))
diff --git a/docker-compose.prod.yml b/docker-compose.prod.yml
index 6a3ceaa82..8fce11645 100644
--- a/docker-compose.prod.yml
+++ b/docker-compose.prod.yml
@@ -46,6 +46,7 @@ services:
       - EMAIL_PROVIDER=
       - EMAIL_SENDER=
       - QUERY_LIST_SORTABLE=true
+      - QUEPID_PROPHET_ANALYTICS=true
     command: "foreman s -f Procfile"
     ports:
     #  - 80:3000 # Map to port 80 for outside users when you are not using Nginx
diff --git a/docs/operating_documentation.md b/docs/operating_documentation.md
index afa6deb1a..abe5adb33 100644
--- a/docs/operating_documentation.md
+++ b/docs/operating_documentation.md
@@ -9,6 +9,7 @@ This document explains how Quepid can be operated and configured.
 - [Legal Pages & GDPR](#legal-pages-&-gdpr)
 - [User Tracking](#user-tracking)
 - [Heathcheck Endpoint](#healthcheck)
+- [Analytics Settings](#analytics-settings)
 - [Troubleshoot Your Deploy](#troubleshoot-your-deploy)
 - [Database Management](#database-management)
 - [Jupyterlite Notebooks](#jupyterlite-notebooks)
@@ -174,6 +175,17 @@ the file `Procfile`
 
 Want to monitor if Quepid is behaving?  Just monitor `/healthcheck`, and you will get 200 status codes from a healthy Quepid, and 503 if not.  The JSON output is `{"code":200,"status":{"database":"OK","migrations":"OK"}}`.
 
+## Analytics Settings
+
+We use the [Prophet.rb](https://github.com/ankane/prophet-ruby) library to decide when interesting things happen in our case scores on the homepage.
+This library may use too much memory for your deploy, and can be disabled.
+
+You can disable this behavior by setting the follow `ENV` var:
+
+```
+QUEPID_PROPHET_ANALYTICS=false
+```
+
 ## Troubleshoot Your Deploy
 
 When errors occur, Quepid logs them and shows a generic page.