I extracted a post call survey voice comment via API and saw this!
,"transcript":" Yeah, the agent are he really helped me out with whatever needed to be sorted out. Thank you.","tags":null}],"channels":1,"provider":"whisper","language":"en-AU","tags":null,"duration":8}
provider = whisper... as in OpenAI Whisper!
It's happening!
To any of the PMs on this board, it would be nice to be able to view / download transcribed comments in bulk to share with the team and celebrate positive feedback.
------------------------------
Ben Fischer
------------------------------
Original Message:
Sent: 12-07-2023 01:40
From: David Maxelon
Subject: Let's Talk Quality Management
not too much to add really,
- The permissions are a bit too complicated and applying them is only intermittently successful
- There should be a field/dashboard for line manager as well as supervisor, supervisors/trainers do the QM but managers need to monitor their teams
- Reporting has been unreliable, our managers have found several errors which have had to be fixed - the product needs a full round of its own quality testing IMO
- Should be able to schedule reports same as the other products
- Share links simply don't work; when you find an issue and want to share with another manager with full admin rights it still doesn't let you access the interaction from the link so the other party has to look it up from scratch.
- Speech analytics - this is useless to us as each department can't put a queue filter in the top level results; we share system with internal and external customers so can't even use this as voice of the customer
- Transcription - it's really bad; I'm told they're changing it over but it can't come soon enough
I'd also like it to be a bit easier to skim through batches of calls (maybe with a read/unread type indicator) so that when trying to trace a rogue call from a page of results you can easily see which ones you've checked.
Feature requests
- Flags AI generated for quality of customer experience
- AI generated three word summary of each call showing in the search page
- There should be a way to use the speech analytics to complete a draft scorecard, automatically flagging calls of concern (e.g. with bad language) and outstanding calls for supervisors to complete full QM on
Sorry I can't help with your issue or suggest any tips other than to suggest an alternative supplier ;-)
------------------------------
David
Original Message:
Sent: 12-04-2023 15:52
From: Ben Fischer
Subject: Let's Talk Quality Management
Firstly, the headaches;
- Main Groups are synced from CM to QM. In some circumstances this doesn't flow through resulting in Agents having Administrator access in QM. Nice.
- There is no way to "resync" the Agents and Groups from CM to QM. Nice
- The default Supervisor group has lots of access. Due to the lack of delegated call monitoring for new agents in VCC, many Agents are granted supervisor access in order to learn from other staff. This results in new hires having access to call recording of senior management. Nice.
- If you choose NOT to use Screen Recording, you can't get rid of the big blue ads for it everywhere in Live Monitoring. Nice.
As for the above, I strongly recommend you disable the supervisor group or make them an Agent to avoid any issues. Also ensure ALL AGENTS have a Main group set. You can generally tell if they don't by the fact they are listed as Administrators on the user list.
The Evaluation reporting that our Operations team does not work out of the box. We have 3 options; which are all great.
Average Quality Scores
Quality Scoring Trends
Evaluation Summary
The Evaluation template created has
Section 1
Question
Question
Question
Section 2
Question
Question
etc
The Operations request was to report by Section, but the Trends only reports by total or question but not Section - and Eval Summary can only dump the data in a spreadsheet, sure it can be manipulated but lacks the more granular options like per Section.
I have having a massive issue with "Supervisor" roles not being able to see ANY interactions atm [even through they have the authorised groups, view interactions and evaluations enabled] which I am trying to work through to determine whether it's a bug or a user issue so will update a bit later.
BUT...
Who else successfully uses Evaluations? Any cool hints or tips to share?
------------------------------
Ben Fischer
------------------------------