diff --git a/.github/workflows/performance-tests-scheduled.yml b/.github/workflows/performance-tests-scheduled.yml
index f326fab753c..195a7c81ed7 100644
--- a/.github/workflows/performance-tests-scheduled.yml
+++ b/.github/workflows/performance-tests-scheduled.yml
@@ -38,6 +38,9 @@ jobs:
             env: |
               DHIS2_IMAGE=dhis2/core-dev:latest
               SIMULATION_CLASS=org.hisp.dhis.test.platform.UsersPerformanceTest
+              DB_DIR=dev
+              DB_TYPE=platform-perf
+              DB_VERSION=44-2026-06-03
               WARMUP=2
               MVN_ARGS="-Diterations=10 -Dprofile=smoke"
             slack_webhook_secret: SLACK_WEBHOOK_PLATFORM_PERFORMANCE
@@ -45,6 +48,9 @@ jobs:
             env: |
               DHIS2_IMAGE=dhis2/core-dev:latest
               SIMULATION_CLASS=org.hisp.dhis.test.platform.UsersPerformanceTest
+              DB_DIR=dev
+              DB_TYPE=platform-perf
+              DB_VERSION=44-2026-06-03
               MVN_ARGS=-Diterations=30
             slack_webhook_secret: SLACK_WEBHOOK_PLATFORM_PERFORMANCE
     uses: ./.github/workflows/performance-tests.yml
diff --git a/dhis-2/dhis-test-performance/src/test/java/org/hisp/dhis/test/platform/UsersPerformanceTest.java b/dhis-2/dhis-test-performance/src/test/java/org/hisp/dhis/test/platform/UsersPerformanceTest.java
index 94b7c819cba..1500a3d020e 100644
--- a/dhis-2/dhis-test-performance/src/test/java/org/hisp/dhis/test/platform/UsersPerformanceTest.java
+++ b/dhis-2/dhis-test-performance/src/test/java/org/hisp/dhis/test/platform/UsersPerformanceTest.java
@@ -55,7 +55,8 @@
 /**
  * Performance test for single-user CRUD operations on {@code /api/users}.
  *
- * <p>Five scenarios, all running against the Sierra Leone demo DB by default:
+ * <p>Five scenarios, all running against the platform-perf DB (~250k users / ~250k org units) by
+ * default:
  *
  * <ol>
  *   <li><b>POST</b> — creates a new user (with optional group assignment)
@@ -71,7 +72,7 @@
  * itself — no existing database users are touched.
  *
  * <p>Run with {@code -DuserGroupUid=<uid>} pointing at a group with large membership to expose N+1
- * problems in POST and DELETE. The default points at "Administrators" on the SL demo DB.
+ * problems in POST and DELETE. The default points at a ~83k-member group on the platform-perf DB.
  *
  * <p>Configuration can be provided via a {@code .properties} file instead of individual {@code -D}
  * flags:
@@ -82,16 +83,16 @@
  *
  * Individual {@code -D} flags always override values from the config file.
  *
- * <p>Available properties (with SL demo DB defaults):
+ * <p>Available properties (with platform-perf DB defaults):
  *
  * <ul>
  *   <li>{@code configFile} — path to a {@code .properties} file (optional)
  *   <li>{@code baseUrl} (default: {@code http://localhost:8080})
  *   <li>{@code username} (default: {@code admin})
  *   <li>{@code password} (default: {@code district})
- *   <li>{@code userRoleUid} (default: {@code Euq3XfEIEbx} — "Data entry clerk")
- *   <li>{@code orgUnitUid} (default: {@code ImspTQPwCqd} — "Sierra Leone" root)
- *   <li>{@code userGroupUid} (default: {@code wl5cDMuUhmF} — "Administrators")
+ *   <li>{@code userRoleUid} (default: {@code MoRvPzDH7lc} — generic role with ~83k users)
+ *   <li>{@code orgUnitUid} (default: {@code VCCdfC9pvMA} — root org unit)
+ *   <li>{@code userGroupUid} (default: {@code KOvR9SAEeEZ} — group with ~83k members)
  *   <li>{@code iterations} (default: {@code 3})
  *   <li>{@code mode} (default: {@code sequential}; runs each scenario in isolation so timings
  *       reflect single-endpoint latency. Use {@code parallel} to run all scenarios concurrently as