diff --git a/utils/radar+/clean_datalake.R b/utils/radar+/clean_datalake.R
new file mode 100644
index 0000000..7514d16
--- /dev/null
+++ b/utils/radar+/clean_datalake.R
@@ -0,0 +1,138 @@
+credentials <<- hublot::get_credentials(
+  Sys.getenv("HUB3_URL"), 
+  Sys.getenv("HUB3_USERNAME"), 
+  Sys.getenv("HUB3_PASSWORD")
+)
+
+medias_urls <- list(
+   cbcnews = list(
+     long_name  = "CBC News",
+     short_name = "CBC",
+     country    = "CAN",
+     base       = "https://www.cbc.ca",
+     front      = "/news"
+   ),
+   jdm = list(
+     long_name  = "Le Journal de Montréal",
+     short_name = "JDM",
+     country    = "CAN",
+     base  = "https://www.journaldemontreal.com",
+     front = "/"
+   ),
+   radiocan = list(
+     long_name  = "Radio-Canada Info",
+     short_name = "RCI",
+     country    = "CAN",
+     base  = "https://ici.radio-canada.ca",
+     front = "/info"
+   ),
+  nationalPost = list(
+    long_name  = "National Post",
+    short_name = "NP",
+    country    = "CAN",
+    base  = "https://nationalpost.com",
+    front = "/"
+  ),
+  tvaNouvelles = list(
+    long_name  = "TVA Nouvelles",
+    short_name = "TVA",
+    country    = "CAN",
+    base  = "https://www.tvanouvelles.ca",
+    front = "/"
+  ),
+  globeAndMail = list(
+    long_name  = "The Globe and Mail",
+    short_name = "GAM",
+    country    = "CAN",
+    base  = "https://www.theglobeandmail.com",
+    front = "/"
+  ),
+  vancouverSun = list(
+    long_name  = "Vancouver Sun",
+    short_name = "VS",
+    country    = "CAN",
+    base  = "https://vancouversun.com",
+    front = "/"
+  ),
+  laPresse = list(
+    long_name  = "La Presse",
+    short_name = "LAP",
+    country    = "CAN",
+    base  = "https://www.lapresse.ca",
+    front = "/"
+  ),
+  leDevoir = list(
+    long_name  = "Le Devoir",
+    short_name = "LED",
+    country    = "CAN",
+    base  = "https://www.ledevoir.com",
+    front = "/"
+  ),
+  montrealGazette = list(
+    long_name  = "Montreal Gazette",
+    short_name = "MG",
+    country    = "CAN",
+    base  = "https://montrealgazette.com",
+    front = "/"
+  ),
+  CTVNews = list(
+    long_name  = "CTV News",
+    short_name = "CTV",
+    country    = "CAN",
+    base  = "https://www.ctvnews.ca",
+    front = "/"
+  ),
+  globalNews = list(
+    long_name  = "Global News",
+    short_name = "GN",
+    country    = "CAN",
+    base  = "https://globalnews.ca",
+    front = "/"
+  ),
+  theStar = list(
+    long_name  = "The Toronto Star",
+    short_name = "TTS",
+    country    = "CAN",
+    base  = "https://www.thestar.com",
+    front = "/"
+  )
+)
+
+datalake_base_path <<- "radarplus"
+
+  r <- hublot::filter_lake_items(
+    filter = list(
+      path=paste(datalake_base_path, "frontpage", sep="/"),
+      metadata__schema = "prod",
+      metadata__start_timestamp__isnull=FALSE
+    ),
+    credentials = credentials
+  )
+
+  length(r$results)
+
+  for (lake_item in r$results) {
+    cat(lake_item$key, " ", lake_item$metadata$schema, " ", lake_item$metadata$start_timestamp, "\n")
+
+    #hublot::remove_lake_item(lake_item$id, credentials)
+  }
+
+
+  r <- hublot::filter_lake_items(
+    filter = list(
+      path=paste(datalake_base_path, "headline", sep="/"),
+      metadata__schema = "prod",
+      metadata__start_timestamp__isnull=FALSE
+    ),
+
+    credentials = credentials
+  )
+
+  length(r$results)
+
+  for (lake_item in r$results) {
+    cat(lake_item$key, " ", lake_item$metadata$schema, " ", lake_item$metadata$start_timestamp, "\n")
+
+    #hublot::remove_lake_item(lake_item$id, credentials)
+  }
+