Vor Kurzem aufgerufene Suchen
Keine vor kurzem aufgerufene Suchen

Allen Lai | Head of CX at Otter.ai
Beigetreten 15. Apr. 2021
·
Letzte Aktivität 10. Dez. 2024
Folge ich
0
Follower
3
Gesamtaktivitäten
150
Stimmen
25
Abonnements
54
AKTIVITÄTSÜBERSICHT
BADGES
BEITRÄGE
POSTS
COMMUNITY-KOMMENTARE
BEITRAGSKOMMENTARE
AKTIVITÄTSÜBERSICHT
Neueste Aktivität von Allen Lai | Head of CX at Otter.ai
Allen Lai | Head of CX at Otter.ai hat einen Kommentar hinterlassen
Gareth Elsby Sure here's how I did it.
-
Create a trigger in Zendesk with the following conditions:
- Meet ALL of the following conditions
- Ticket status Changed from Solved
- Ticket status Changed to Open
- Actions
- Add tags “reopen_solved”
- Meet ALL of the following conditions
-
Create a custom view in Zendesk with the following conditions:
- Tickets must meet all of these conditions to appear in the view
- Ticket status contains at least one of the following: Open
- Hours since status category open (calendar) is 1
- Tags contains at least one of the following: reopen_solved
- Tickets must meet all of these conditions to appear in the view
-
In Zapier, do the following:
-
Step 1: Zendesk
- Event: New Ticket in View
- View: Select the view created in the previous step
-
Step 2: Delay by Zapier
- Add a one-minute delay to ensure the trigger in step 1 occurs first.
-
Step 3: Filter by Zapier
- Only continue if:
- Status Contains open; AND
- Tags Contains reopen_solved
- Only continue if:
-
Step 4: Zendesk
- Event: Find Latest Comment
-
Step 5: Filter by Zapier
- Only continue if:
- From Step 4, “Plain Body” Contains “thank you”; AND
- From Step 4, “Via Source From Address” Exists; OR
- From Step 4, “Plain Body” Contains “thanks”; AND
- From Step 4, “Via Source From Address” Exists
- Only continue if:
-
Step 6: OpenAI
- Event: Send Prompt
- Model: gpt-3.5-instruct
- Prompt: "Review the first 255 characters of the message below from a customer. If the message is purely an expression of thanks without any indication that a reply or follow-up is needed, respond with 'TRUE'. Otherwise, respond with 'FALSE'. Note that any mention of an issue or request for assistance should lead to a 'FALSE' response. Here is the customer's message:
"4. Plain Body"
-
Step 7: Filter by Zapier
- Only continue if:
- From Step 6, Response Contains “true”
- Only continue if:
-
Step 8: Zendesk
- Event: Update Ticket
- New status: Solved
- New Comment to add to the Ticket: “Zapier automatically updated the ticket status from Open to Solved because the customer was simply thanking us and a reply or follow-up is not needed.”
- Should the new Comment be public? No
-
Step 1: Zendesk
Kommentar anzeigen · Bearbeitet 10. Dez. 2024 · Allen Lai | Head of CX at Otter.ai
0
Follower
0
Stimmen
0
Kommentare
Allen Lai | Head of CX at Otter.ai hat einen Kommentar hinterlassen
I would like to be able to create Views based on the high-level intent, such as Account, Billing, etc. Currently, I can only create Views based on a specific intent, which is very time-consuming.
Kommentar anzeigen · Gepostet 21. Nov. 2024 · Allen Lai | Head of CX at Otter.ai
0
Follower
0
Stimmen
0
Kommentare
Allen Lai | Head of CX at Otter.ai hat einen Kommentar hinterlassen
We're thinking of implementing an internal KB, and my only concern is that agents accidentally link to an internal article within their public response. Will it allow agents to do that?
Kommentar anzeigen · Gepostet 23. Okt. 2024 · Allen Lai | Head of CX at Otter.ai
0
Follower
0
Stimmen
0
Kommentare
Allen Lai | Head of CX at Otter.ai hat einen Post erstellt
I'm looking to create a report showing the average number of agent comments required to resolve a ticket, but I'm having trouble finding an obvious way to do this. Has anyone else had success with this?
Gepostet 21. Aug. 2024 · Allen Lai | Head of CX at Otter.ai
0
Follower
0
Stimmen
0
Kommentare
Allen Lai | Head of CX at Otter.ai hat einen Post erstellt
I notice that when an agent puts themselves on “Unavailable” using the Out of Office app, it adds the tag “agent_ooo” to their tickets. When they return from vacation and set themselves back to “Available,” it removes the “agent_ooo” tag. When I look at a report that shows how many tickets they've updated, this skews their numbers.
I was hoping to create a standard calculated attribute that I could filter on, but I can't quite figure if it's possible. Something like the following is what I'm looking for:
IF ([Changes - Field name] = "tags" AND
CONTAINS [Changes - New value], "agent_ooo")
THEN "true"
ENDIF
Similarly, I'd need another to filter when the tag is removed.
Gepostet 08. Aug. 2024 · Allen Lai | Head of CX at Otter.ai
0
Follower
0
Stimmen
0
Kommentare
Allen Lai | Head of CX at Otter.ai hat einen Kommentar hinterlassen
When an email is forwarded to our support address, is there any attribute I can use within a trigger to tag these tickets? I want to differentiate these tickets by adding a tag or updating one of the ticket fields. Thanks!
Kommentar anzeigen · Gepostet 31. Mai 2024 · Allen Lai | Head of CX at Otter.ai
0
Follower
0
Stimmen
0
Kommentare
Allen Lai | Head of CX at Otter.ai hat einen Kommentar hinterlassen
I created an automation via Zapier using a prompt in OpenAI, which is working wonders for us. I would highly recommend it.
Kommentar anzeigen · Gepostet 04. Mai 2024 · Allen Lai | Head of CX at Otter.ai
0
Follower
1
Stimme
0
Kommentare
Allen Lai | Head of CX at Otter.ai hat einen Kommentar hinterlassen
From my understanding, the Requester Wait Time is only calculated after a ticket is updated. So you don’t get an accurate way to gauge the average wait time of all New and Open tickets. I learned this the hard way when I manually exported our tickets and then calculated it myself, and found that the value for Requester Wait Time was severely understated. Seems like something that should be improved, otherwise, you’re reporting inaccurate information.
Kommentar anzeigen · Gepostet 02. Mai 2024 · Allen Lai | Head of CX at Otter.ai
0
Follower
1
Stimme
0
Kommentare
Allen Lai | Head of CX at Otter.ai hat einen Kommentar hinterlassen
I'm trying to calculate the Requester Wait Time to include the time it spends in New or Open status before an agent has replied. We sometimes have tickets in Open status if automation attempts to answer the customer's request but is unsuccessful.
Here's my attempt, but I wasn't sure if this was the right way to calculate it.
IF ([Ticket status]="New") OR (([Ticket status]="Open" AND [Assignee stations brackets]="NULL" AND VALUE(Agent replies)<1))
THEN DATE_DIFF(NOW(),[Ticket created - Timestamp],"nb_of_hours")
ELSE VALUE(Requester wait time - Business hours (min))/60
ENDIF
Kommentar anzeigen · Gepostet 24. Apr. 2024 · Allen Lai | Head of CX at Otter.ai
0
Follower
0
Stimmen
0
Kommentare
Allen Lai | Head of CX at Otter.ai hat einen Kommentar hinterlassen
Amy Dee Do you know a way to calculate the 90th percentile of the Requester wait time?
Kommentar anzeigen · Gepostet 23. Apr. 2024 · Allen Lai | Head of CX at Otter.ai
0
Follower
0
Stimmen
0
Kommentare