Sales ops · Production

How Siro reduced customer complaints and support tickets by 90% with AssemblyAI

The problem

Field sales representatives across Siro's customer base lacked adequate coaching because in-person coaching required too much time and capital for companies with dispersed teams, reducing company oversight and limiting rep earnings.

First attempt

Siro's initial legacy speech-to-text provider produced inaccurate transcriptions, causing support tickets to build up and customers to churn before they switched to AssemblyAI.

Workflow diagram · grounded in source
1
Field call recorded on mobile
Trigger
Sales agents record customer interactions in the field on mobile devices such as cell phones or tablets.
▾ source quote
“recordings on mobile devices (e.g., cell phones or tablets)”
2
Speech transcription
Ai action
AssemblyAI's Universal speech recognition model transcribes the field recording with best-in-class accuracy.
▾ source quote
“AssemblyAI's Universal speech recognition model offers best-in-class transcription and speaker diarization, even in noisy, less-than-ideal recording environments, as well as low latency at 23 seconds on a 30-minute audio file”
3
Speaker diarization
Ai action
Speaker identification and labeling automatically separates sales agents from customers in the recording.
▾ source quote
“Highly accurate speaker diarization, or speaker identification and labeling, to automatically separate sales agents and customers on these recordings”
4
Coaching insights generated
Ai action
Siro generates coaching insights as the core of its platform, dependent on accurate transcription and speaker diarization.
▾ source quote
“the coaching insights Siro generates—the core of the Siro platform—would fall apart”
5
Coaching delivered to reps
Output
Sales reps receive coaching 10x faster through the Siro platform.
▾ source quote
“Receive coaching 10x faster”
Reported outcome

With AssemblyAI, Siro reduced customer complaints and support tickets by 90%, and field sales reps using the platform see a 20-40% increase in sales growth, a 36% improvement in close rate, and 10x faster coaching delivery.

Reported metrics
Customer complaints and support tickets90%
Sales growth increase20-40%
Close rate improvement36%
Ride-along time savings10x
Show all 9 reported metrics
customer complaints and support tickets90%
sales growth increase20-40%
close rate improvement36%
ride-along time savings10x
coaching speed improvement10x faster
rep earnings per quarter increase$20,000 per quarter
speaker diarization model accuracy improvement13%
speaker count errors reduction85.4%
transcription latency23 seconds on a 30-minute audio file
◆ Does this fit your context?

Compare to your context

Tell us your scale, team, and constraints. We'll show what changes at your size, what fails at your scale, and whether this case is a fit, needs adaptation, or won't scale to you. Free demo, no signup.

Compare to your context →
~30 seconds · free
Source
https://www.assemblyai.com/customers/siro-customer-story
Read source ↗

Frequently asked questions

What did this team achieve with this AI workflow?

With AssemblyAI, Siro reduced customer complaints and support tickets by 90%, and field sales reps using the platform see a 20-40% increase in sales growth, a 36% improvement in close rate, and 10x faster coaching del…

What results were reported?

Customer complaints and support tickets: 90%; Sales growth increase: 20-40%; Close rate improvement: 36%; Ride-along time savings: 10x (source-reported, not independently verified).

What failed first in this deployment?

Siro's initial legacy speech-to-text provider produced inaccurate transcriptions, causing support tickets to build up and customers to churn before they switched to AssemblyAI.

How is this sales ops AI workflow structured?

Field call recorded on mobile → Speech transcription → Speaker diarization → Coaching insights generated → Coaching delivered to reps.

Related sales ops cases

WHAT TO DO WITH THIS

Now compare it to your context

This case is one data point. Whether its pattern fits you depends on your volumes, your stack, and your exception load — that comparison is the step no case study can do for you.

Browse the Sales ops cases →