The Night Emails Died: AI Cleanup Incident Explained

1
00:00:00,000 –> 00:00:02,000
The city got quiet.

2
00:00:02,000 –> 00:00:03,720
Too quiet.

3
00:00:03,720 –> 00:00:06,680
You know that feeling when you walk into an office on a Monday morning,

4
00:00:06,680 –> 00:00:09,640
expecting the usual chaos, the phones ringing off the hook,

5
00:00:09,640 –> 00:00:11,680
the inbox screaming at you.

6
00:00:11,680 –> 00:00:15,240
But instead, it’s just silence.

7
00:00:15,240 –> 00:00:18,400
I do. It’s unnerving. It feels like the calm before the storm,

8
00:00:18,400 –> 00:00:20,200
or in the world of customer service.

9
00:00:20,200 –> 00:00:21,960
It usually means the server is down.

10
00:00:21,960 –> 00:00:23,520
Exactly.

11
00:00:23,520 –> 00:00:25,480
But in the story we’re looking at today,

12
00:00:25,480 –> 00:00:29,160
the silence wasn’t a crash. It was a cleanup.

13
00:00:29,160 –> 00:00:33,560
We’re diving into a narrative that frames modern customer service automation,

14
00:00:33,560 –> 00:00:38,800
specifically using Dynamics 365 as a noir detective story.

15
00:00:38,800 –> 00:00:40,960
It’s gritty, it’s dramatic,

16
00:00:40,960 –> 00:00:44,920
and it’s surprisingly accurate about the mess most companies are in.

17
00:00:44,920 –> 00:00:46,160
It’s a brilliant metaphor,

18
00:00:46,160 –> 00:00:48,440
because if you think about a shared inbox today,

19
00:00:48,440 –> 00:00:49,840
it really is a crime scene.

20
00:00:49,840 –> 00:00:51,120
Deadletters everywhere,

21
00:00:51,120 –> 00:00:53,040
customers screaming into the void,

22
00:00:53,040 –> 00:00:56,280
cases rotting like, well, let’s stick to the noir theme,

23
00:00:56,280 –> 00:00:58,560
like forgotten bodies in an alleyway.

24
00:00:58,560 –> 00:01:00,360
That is a vivid image,

25
00:01:00,360 –> 00:01:03,360
and frankly, a bit gross, but it hits home.

26
00:01:03,360 –> 00:01:07,360
We’ve all seen that shared mailbox where emails go to die.

27
00:01:07,360 –> 00:01:08,840
Deadletters, that’s the phrase,

28
00:01:08,840 –> 00:01:10,880
and the crime isn’t that people aren’t working hard.

29
00:01:10,880 –> 00:01:12,920
The crime is the system. It’s manual triage.

30
00:01:12,920 –> 00:01:15,080
It’s what I call rooting by vibe.

31
00:01:15,080 –> 00:01:16,200
Routing by vibe?

32
00:01:16,200 –> 00:01:18,000
Yeah, you know, I like billing questions,

33
00:01:18,000 –> 00:01:20,120
so I’ll take this one or I’m tired,

34
00:01:20,120 –> 00:01:23,640
so I’ll leave that complex technical issue for someone else.

35
00:01:23,640 –> 00:01:26,440
It’s chaotic. It’s based on human mood, not business logic,

36
00:01:26,440 –> 00:01:29,000
and that’s where the night the emails died comes in.

37
00:01:29,000 –> 00:01:31,480
It introduces three specific autonomous agents,

38
00:01:31,480 –> 00:01:33,720
the cleanup crew that solve this crime.

39
00:01:33,720 –> 00:01:36,320
I love that, the cleanup crew.

40
00:01:36,320 –> 00:01:38,880
So let’s walk through this crime scene.

41
00:01:38,880 –> 00:01:42,480
We have a victim, the customer experience.

42
00:01:42,480 –> 00:01:45,720
We have a suspect, the legacy inbox.

43
00:01:45,720 –> 00:01:49,240
Who is the first operator that steps in to clean up the streets?

44
00:01:49,240 –> 00:01:51,360
The first operator is the case scanner.

45
00:01:51,360 –> 00:01:54,320
Think of it as the detective with the camera at the crime scene.

46
00:01:54,320 –> 00:01:56,480
In a traditional setup, an email comes in,

47
00:01:56,480 –> 00:01:58,400
and a human has to open it, read it,

48
00:01:58,400 –> 00:02:00,520
figure out if it’s angry or happy.

49
00:02:00,520 –> 00:02:03,440
Look for an order number, maybe download an attachment.

50
00:02:03,440 –> 00:02:04,400
It’s slow.

51
00:02:04,400 –> 00:02:06,320
And humans hesitate.

52
00:02:06,320 –> 00:02:08,680
We skim. We miss things.

53
00:02:08,680 –> 00:02:10,600
Precisely. The case scanner doesn’t blink.

54
00:02:10,600 –> 00:02:13,280
It’s using email to case ingestion, but on steroids.

55
00:02:13,280 –> 00:02:14,680
It’s not just forwarding the email.

56
00:02:14,680 –> 00:02:15,800
It’s stripping it for parts.

57
00:02:15,800 –> 00:02:19,160
It reads the subject line, the body text, even the footer.

58
00:02:19,160 –> 00:02:20,400
But here is the kicker.

59
00:02:20,400 –> 00:02:22,240
It reads the attachments, too.

60
00:02:22,240 –> 00:02:26,320
Wait, so if I send a screenshot of a broken product or a PDF receipt,

61
00:02:26,320 –> 00:02:28,560
the scanner is actually analyzing that image

62
00:02:28,560 –> 00:02:30,640
before a human ever sees it.

63
00:02:30,640 –> 00:02:33,720
Yes, OCR, optical character recognition.

64
00:02:33,720 –> 00:02:36,080
It sees a photo of a jacket with a split zipper,

65
00:02:36,080 –> 00:02:37,960
tags it as damaged goods.

66
00:02:37,960 –> 00:02:41,160
It sees a PDF contract, extracts the policy number.

67
00:02:41,160 –> 00:02:44,720
It turns unstructured noise into structured evidence.

68
00:02:44,720 –> 00:02:47,240
It creates the case file, fills in the fields,

69
00:02:47,240 –> 00:02:50,200
customer product priority, and stamps it.

70
00:02:50,200 –> 00:02:51,840
That changes the game.

71
00:02:51,840 –> 00:02:53,960
You’re not starting from zero anymore.

72
00:02:53,960 –> 00:02:56,080
You’re starting with a file that’s already built.

73
00:02:56,080 –> 00:02:59,600
Exactly. The dead letter is revived before it even hits the floor.

74
00:02:59,600 –> 00:03:01,360
But that leads us to the second problem.

75
00:03:01,360 –> 00:03:02,720
You have a file, but who solves it?

76
00:03:02,720 –> 00:03:04,720
In the old city, you just shout into the room,

77
00:03:04,720 –> 00:03:06,080
who handles returns.

78
00:03:06,080 –> 00:03:10,920
Right. Or it sits in a general queue until someone cherry picks it.

79
00:03:10,920 –> 00:03:13,720
Which brings us to the second operator, the traffic controller.

80
00:03:13,720 –> 00:03:15,520
This is unified routing, the grid.

81
00:03:15,520 –> 00:03:19,760
This sounds less like a detective and more like air traffic control.

82
00:03:19,760 –> 00:03:20,960
That’s a fair comparison.

83
00:03:20,960 –> 00:03:23,520
The traffic controller stands over the map of the city.

84
00:03:23,520 –> 00:03:24,880
It doesn’t care about vibes.

85
00:03:24,880 –> 00:03:26,360
It cares about three things.

86
00:03:26,360 –> 00:03:28,640
Rules, skills, and capacity.

87
00:03:28,640 –> 00:03:31,200
It looks at that case, the scanner just built, say,

88
00:03:31,200 –> 00:03:34,240
a high priority return for a VIP customer.

89
00:03:34,240 –> 00:03:36,040
And it looks at the workforce.

90
00:03:36,040 –> 00:03:38,680
So it knows that agent Rivera is good at returns,

91
00:03:38,680 –> 00:03:41,440
but agent Smith is better at technical support.

92
00:03:41,440 –> 00:03:42,440
It goes deeper.

93
00:03:42,440 –> 00:03:45,920
It knows agent Rivera is good at returns, speaks Spanish,

94
00:03:45,920 –> 00:03:48,400
and currently has capacity for one more case.

95
00:03:48,400 –> 00:03:50,480
It knows agent Smith is technically capable

96
00:03:50,480 –> 00:03:52,840
but is already redlining on three other tickets.

97
00:03:52,840 –> 00:03:55,440
It routes the case like a light through an intersection,

98
00:03:55,440 –> 00:03:56,840
no wandering souls.

99
00:03:56,840 –> 00:04:00,000
That eliminates the cherry picking problem entirely.

100
00:04:00,000 –> 00:04:00,840
Completely.

101
00:04:00,840 –> 00:04:02,160
And it prevents burnout.

102
00:04:02,160 –> 00:04:04,560
If the lane is clogged, the controller holds the light red

103
00:04:04,560 –> 00:04:05,800
or opens a new lane.

104
00:04:05,800 –> 00:04:06,800
It’s dynamic.

105
00:04:06,800 –> 00:04:09,040
And the best part, it keeps receipts.

106
00:04:09,040 –> 00:04:10,840
If a case goes to the wrong person,

107
00:04:10,840 –> 00:04:13,360
you have a flight recorder, diagnostic.

108
00:04:13,360 –> 00:04:17,400
You can see exactly which rule sent it there and fix the rule.

109
00:04:17,400 –> 00:04:19,760
You don’t blame the person, you fix the logic.

110
00:04:19,760 –> 00:04:21,760
That’s a huge cultural shift.

111
00:04:21,760 –> 00:04:24,000
You stop asking, why did you take this?

112
00:04:24,000 –> 00:04:26,800
And start asking, why did the system send this?

113
00:04:26,800 –> 00:04:28,600
It takes the politics out of the queue.

114
00:04:28,600 –> 00:04:32,480
It cleans up the streets, but we still have one massive bottleneck lift.

115
00:04:32,480 –> 00:04:35,400
The case is created, it’s routed to the right person.

116
00:04:35,400 –> 00:04:38,160
But now that person has to actually write the response.

117
00:04:38,160 –> 00:04:40,280
The dreaded blinking cursor.

118
00:04:40,280 –> 00:04:43,760
The, thank you for your email, we value your business.

119
00:04:43,760 –> 00:04:46,160
Typing that out a hundred times a day.

120
00:04:46,160 –> 00:04:48,120
It’s soul crushing and it’s slow.

121
00:04:48,120 –> 00:04:49,440
Enter the third operator.

122
00:04:49,440 –> 00:04:50,440
The shadow operator.

123
00:04:50,440 –> 00:04:51,880
That sounds ominous.

124
00:04:51,880 –> 00:04:53,960
It’s actually the most helpful partner you could ask for.

125
00:04:53,960 –> 00:04:55,440
This is co-pilot studio.

126
00:04:55,440 –> 00:04:58,120
It sits in the room wired into the service.

127
00:04:58,120 –> 00:05:00,440
While the traffic controller is rooting the case,

128
00:05:00,440 –> 00:05:02,720
the shadow operator is already reading it.

129
00:05:02,720 –> 00:05:05,120
It scans the archive, checks the knowledge base,

130
00:05:05,120 –> 00:05:06,760
and drafts the reply.

131
00:05:06,760 –> 00:05:10,200
So when the agent opens the case, the answer is already there?

132
00:05:10,200 –> 00:05:11,280
Not just an answer.

133
00:05:11,280 –> 00:05:11,880
The answer.

134
00:05:11,880 –> 00:05:12,800
It doesn’t ask questions.

135
00:05:12,800 –> 00:05:14,760
The company already knows the answers to.

136
00:05:14,760 –> 00:05:18,280
You know how frustrating it is when a company asks for your order number

137
00:05:18,280 –> 00:05:19,920
when it was in the subject line?

138
00:05:19,920 –> 00:05:21,840
Oh, it drives me crazy.

139
00:05:21,840 –> 00:05:23,960
It feels like they aren’t listening.

140
00:05:23,960 –> 00:05:25,560
The shadow operator listens.

141
00:05:25,560 –> 00:05:27,160
It sees the order number.

142
00:05:27,160 –> 00:05:28,960
It sees the policy on returns.

143
00:05:28,960 –> 00:05:32,920
It drafts a response that says, “Hi, I see your zipper split on the forest green jacket.

144
00:05:32,920 –> 00:05:34,400
I’ve initiated a replacement.

145
00:05:34,400 –> 00:05:36,000
Here is your return label.”

146
00:05:36,000 –> 00:05:41,520
It cites the specific knowledge base article, say KB2499 and presents it to the human agent.

147
00:05:41,520 –> 00:05:43,760
So the human agent isn’t the writer anymore.

148
00:05:43,760 –> 00:05:45,000
They’re the editor.

149
00:05:45,000 –> 00:05:46,000
Exactly.

150
00:05:46,000 –> 00:05:46,840
They’re the judge.

151
00:05:46,840 –> 00:05:48,080
The shadow speaks.

152
00:05:48,080 –> 00:05:49,960
But the human pulls the trigger.

153
00:05:49,960 –> 00:05:55,680
The agent reviews the draft, maybe softens the tone, maybe checks the logic and hits send.

154
00:05:55,680 –> 00:06:00,960
What used to take 10 minutes of hunting for info and typing takes 30 seconds of review?

155
00:06:00,960 –> 00:06:03,720
That phrase from the story really stuck with me.

156
00:06:03,720 –> 00:06:05,520
Faster than regret.

157
00:06:05,520 –> 00:06:06,760
It’s poetic, isn’t it?

158
00:06:06,760 –> 00:06:09,000
It implies that speed isn’t just about efficiency.

159
00:06:09,000 –> 00:06:10,720
It’s about emotional salvage.

160
00:06:10,720 –> 00:06:14,560
If you wait 48 hours to reply to a complaint, the customer has already moved from annoyed

161
00:06:14,560 –> 00:06:15,720
to furious.

162
00:06:15,720 –> 00:06:16,960
You’re managing regret.

163
00:06:16,960 –> 00:06:19,800
If you reply in three minutes with a solution, you’re a hero.

164
00:06:19,800 –> 00:06:23,240
Let’s look at the real case files mentioned in the story.

165
00:06:23,240 –> 00:06:26,400
They broke it down into retail, insurance and HR.

166
00:06:26,400 –> 00:06:31,040
I want to dig into the insurance one because that feels high stakes.

167
00:06:31,040 –> 00:06:34,080
Case number 0228.

168
00:06:34,080 –> 00:06:35,240
The flooded basement.

169
00:06:35,240 –> 00:06:38,280
This is a classic example of severity hiding in plain sight.

170
00:06:38,280 –> 00:06:41,000
In the old system, an email says water all over.

171
00:06:41,000 –> 00:06:45,240
It sits in the pile with “I lost my password” but to the customer, their house is destroying

172
00:06:45,240 –> 00:06:46,240
itself.

173
00:06:46,240 –> 00:06:47,240
Right.

174
00:06:47,240 –> 00:06:50,800
Every minute that water sits there, the claim gets more expensive.

175
00:06:50,800 –> 00:06:54,280
The case scanner reads “standing water, drywall and basement”.

176
00:06:54,280 –> 00:06:56,920
It flags it as “property damage, severe”.

177
00:06:56,920 –> 00:07:00,000
The traffic controller sees this isn’t a job for a junior rep.

178
00:07:00,000 –> 00:07:03,600
It roots it immediately to a property adjuster with flood skills.

179
00:07:03,600 –> 00:07:04,760
And the shadow operator?

180
00:07:04,760 –> 00:07:07,520
It drafts a reply that doesn’t say we received your request.

181
00:07:07,520 –> 00:07:10,320
It says, “We’ve logged claim 8841.

182
00:07:10,320 –> 00:07:12,040
Please send two photos at eye level.

183
00:07:12,040 –> 00:07:13,760
Here is the link to the upload portal.”

184
00:07:13,760 –> 00:07:15,080
It’s immediate action.

185
00:07:15,080 –> 00:07:16,640
It stops the bleeding.

186
00:07:16,640 –> 00:07:18,520
And think about the agent experience there.

187
00:07:18,520 –> 00:07:22,280
They didn’t have to triage 700 emails to find that one emergency.

188
00:07:22,280 –> 00:07:24,520
The system handed it to them on a silver platter.

189
00:07:24,520 –> 00:07:28,480
They stopped playing archaeologist, brushing dust off old files, and started doing their

190
00:07:28,480 –> 00:07:30,720
actual job, which is helping people.

191
00:07:30,720 –> 00:07:33,120
What about the HR example?

192
00:07:33,120 –> 00:07:38,960
That one seemed quieter, less dramatic than a flood, but just as messy.

193
00:07:38,960 –> 00:07:40,920
The black hole of BPO.

194
00:07:40,920 –> 00:07:43,440
Shared inboxes where resumes and contracts vanish.

195
00:07:43,440 –> 00:07:46,200
The story mentions a need help subject line.

196
00:07:46,200 –> 00:07:48,000
In the old world, that’s a mystery.

197
00:07:48,000 –> 00:07:52,120
In the new world, the scanner opens the attachment, sees it’s a contract, detects benefits

198
00:07:52,120 –> 00:07:55,480
enrollment, and roots it to the onboarding specialist.

199
00:07:55,480 –> 00:07:56,880
And the shadow operator?

200
00:07:56,880 –> 00:08:00,280
It pulls the specific “Welcome aboard” steps from the internal wiki.

201
00:08:00,280 –> 00:08:02,520
It stitches the reply from facts, not a template.

202
00:08:02,520 –> 00:08:06,880
It says, “Here are your next three actions, and links the actual forms.

203
00:08:06,880 –> 00:08:10,760
It turns a vague, cry for help into a completed process.”

204
00:08:10,760 –> 00:08:14,120
There’s a philosophical shift here that I find really interesting.

205
00:08:14,120 –> 00:08:18,480
We often talk about AI taking jobs, but this narrative frames it differently.

206
00:08:18,480 –> 00:08:25,000
It says, “Humans are for judgment, negotiation, and edge cases, not for sifting the gutter.”

207
00:08:25,000 –> 00:08:26,000
That is the core message.

208
00:08:26,000 –> 00:08:29,080
We have been asking humans to act like machines for 20 years.

209
00:08:29,080 –> 00:08:31,440
Read this code, copy it here, paste it there.

210
00:08:31,440 –> 00:08:32,600
Humans are bad at that.

211
00:08:32,600 –> 00:08:34,080
We get bored, we get tired.

212
00:08:34,080 –> 00:08:35,440
Machines are excellent at it.

213
00:08:35,440 –> 00:08:39,320
By letting the cleanup crew handle the intake, routing, and drafting, you let the human

214
00:08:39,320 –> 00:08:40,320
be human.

215
00:08:40,320 –> 00:08:41,880
You let them use empathy.

216
00:08:41,880 –> 00:08:44,800
But is there a risk of it becoming too cold?

217
00:08:44,800 –> 00:08:48,440
The story mentions cold hands, steady pulse.

218
00:08:48,440 –> 00:08:52,000
If the AI is drafting everything, do we lose the personal touch?

219
00:08:52,000 –> 00:08:54,880
That’s the noir element, the fear of the cold machine.

220
00:08:54,880 –> 00:08:55,880
But look at the result.

221
00:08:55,880 –> 00:08:58,680
Is it more personal to have a human, write a generic?

222
00:08:58,680 –> 00:09:01,720
We are experiencing high volume email after three days?

223
00:09:01,720 –> 00:09:06,800
Or is it more personal to get an immediate, accurate solution drafted by AI and approved

224
00:09:06,800 –> 00:09:07,800
by a human?

225
00:09:07,800 –> 00:09:09,560
That’s a great point.

226
00:09:09,560 –> 00:09:12,000
The importance is its own form of empathy.

227
00:09:12,000 –> 00:09:15,360
Exactly, respecting my time is the highest form of customer service.

228
00:09:15,360 –> 00:09:17,760
And there’s a governance layer here too.

229
00:09:17,760 –> 00:09:21,760
The story emphasizes that human judgment stays on the trigger.

230
00:09:21,760 –> 00:09:26,560
The shadow operator drafts, it doesn’t send, the scanner tags, it doesn’t delete.

231
00:09:26,560 –> 00:09:28,880
The human remains the sheriff of the city.

232
00:09:28,880 –> 00:09:32,240
So the night the emails died, isn’t a tragedy.

233
00:09:32,240 –> 00:09:34,280
It’s the night the noise died.

234
00:09:34,280 –> 00:09:35,840
It’s the night the clutter died.

235
00:09:35,840 –> 00:09:40,080
The night the anxiety of the unread inbox died, when you clear away the noise, you can actually

236
00:09:40,080 –> 00:09:41,200
hear the customer.

237
00:09:41,200 –> 00:09:44,920
And frankly, for the businesses running these systems, you can finally see the data.

238
00:09:44,920 –> 00:09:45,920
The receipts?

239
00:09:45,920 –> 00:09:46,920
The receipts.

240
00:09:46,920 –> 00:09:48,400
You can’t improve what you can’t measure.

241
00:09:48,400 –> 00:09:52,200
If your inbox is a chaotic pile, you don’t know why customers are churning.

242
00:09:52,200 –> 00:09:55,640
Once you structure it, scanner, controller, shadow, you get logs.

243
00:09:55,640 –> 00:09:59,840
You can see, oh, 30% of our volume is about zippers.

244
00:09:59,840 –> 00:10:01,880
Okay, talk to manufacturing.

245
00:10:01,880 –> 00:10:02,880
Fix the zipper.

246
00:10:02,880 –> 00:10:06,200
Move from fixing the ticket to fixing the root cause.

247
00:10:06,200 –> 00:10:07,200
That is the ultimate goal.

248
00:10:07,200 –> 00:10:08,440
Continuous improvement.

249
00:10:08,440 –> 00:10:11,800
You stop just bailing water out of the boat and you finally plug the hole.

250
00:10:11,800 –> 00:10:14,960
I want to circle back to the noir demo concept.

251
00:10:14,960 –> 00:10:20,360
The story describes the speed of this interaction as three seconds faster than regret.

252
00:10:20,360 –> 00:10:22,520
It’s such a powerful hook.

253
00:10:22,520 –> 00:10:26,720
For a business listening to this, someone who is maybe drowning in their own crime scene

254
00:10:26,720 –> 00:10:29,600
of an inbox, what is the first step?

255
00:10:29,600 –> 00:10:32,320
Do they just turn all three on at once?

256
00:10:32,320 –> 00:10:33,960
You can, but it’s usually a progression.

257
00:10:33,960 –> 00:10:35,600
You start with the scanner.

258
00:10:35,600 –> 00:10:36,600
Stop the bleeding.

259
00:10:36,600 –> 00:10:39,080
Get visibility into what is actually coming in.

260
00:10:39,080 –> 00:10:42,600
Stop treating email as text and start treating it as data.

261
00:10:42,600 –> 00:10:45,280
Once you have structured data, then you turn on the traffic controller.

262
00:10:45,280 –> 00:10:46,320
You build the grid.

263
00:10:46,320 –> 00:10:48,080
You stop routing by vibe.

264
00:10:48,080 –> 00:10:50,640
And finally, the shadow operator.

265
00:10:50,640 –> 00:10:52,720
Once your routing is clean, you empower the agents.

266
00:10:52,720 –> 00:10:53,960
You give them the shadow operator.

267
00:10:53,960 –> 00:10:55,120
It’s the force multiplier.

268
00:10:55,120 –> 00:10:59,200
Suddenly, a team of 10 can do the work of 20, not because they’re working harder,

269
00:10:59,200 –> 00:11:03,440
but because they aren’t wasting time typing best regards 50 times an hour.

270
00:11:03,440 –> 00:11:04,520
It’s compelling.

271
00:11:04,520 –> 00:11:09,400
It turns the support center from a cost center, a place where money goes to die, into a strategic

272
00:11:09,400 –> 00:11:10,400
asset.

273
00:11:10,400 –> 00:11:12,080
And it changes the life of the agent.

274
00:11:12,080 –> 00:11:15,520
I think that’s often the overlooked part we talk about custom experience, but employee

275
00:11:15,520 –> 00:11:17,320
experience matters too.

276
00:11:17,320 –> 00:11:19,400
Nobody wants to work in a crime scene.

277
00:11:19,400 –> 00:11:23,880
Everyone wants to work in a clean, efficient city where they have the tools to solve problems.

278
00:11:23,880 –> 00:11:25,360
The city breathes.

279
00:11:25,360 –> 00:11:26,360
The city breathes.

280
00:11:26,360 –> 00:11:27,800
The panic subsides.

281
00:11:27,800 –> 00:11:32,160
You can finally go home at 5pm, knowing there isn’t a ticking time bomb in the shared mailbox.

282
00:11:32,160 –> 00:11:36,600
So for our listeners, if your inbox still runs your city, if you’re still seeing those

283
00:11:36,600 –> 00:11:41,680
dead letters and feeling that dread on Sunday night, maybe it’s time to call in the clean-up

284
00:11:41,680 –> 00:11:42,680
crew.

285
00:11:42,680 –> 00:11:46,560
The case scanner, the traffic controller and the shadow operator, they’re ready to work.

286
00:11:46,560 –> 00:11:47,560
And they don’t sleep.

287
00:11:47,560 –> 00:11:48,800
No mercy for the backlog.

288
00:11:48,800 –> 00:11:51,640
I think that’s the perfect place to leave it.

289
00:11:51,640 –> 00:11:54,800
The night the emails died, isn’t a horror story.

290
00:11:54,800 –> 00:11:56,960
It’s a success story waiting to happen.

291
00:11:56,960 –> 00:11:59,880
Thanks for breaking down the case files with us today.

292
00:11:59,880 –> 00:12:01,480
Always a pleasure to walk the beat.

293
00:12:01,480 –> 00:12:04,360
Until next time, keep your cues clean and your receipts handy.