Customer support · Production

Borrowell powers up Zendesk ticketing with Lyro AI Agent

The problem

Borrowell's support team was managing more than 11,000 tickets per month with only nine agents, with volume spikes climbing beyond 12,000 tickets per month, and the company anticipated further growth would make scaling support without adding headcount difficult.

Workflow diagram · grounded in source

Lyro engages before ticket creation

trigger

“introduced Tidio's Lyro AI to handle conversations before they turned into tickets”

Lyro answers from help center knowledge

ai_action

“Lyro is trained on Borrowell's Zendesk help center articles”

Manual Q&A training for gaps

ai_action

“Manual training: ability to add specific Q&As for unanswered questions”

Member contact details collected

integration

“Member contact details are collected during the conversation”

Unresolved chats become tickets

routing

“Conversations Lyro can't resolve are automatically converted into Zendesk tickets”

Full transcript attached

integration

“Full chat transcripts are included in every ticket”

Ticket routed to support queue

routing

“Tickets are routed to the appropriate support queues”

Agent reviews with full context

human_review

“agents always have context and can step in without repeating questions”

Reported outcome

Lyro AI Agent achieved an 83% resolution rate, an 87% answer rate, and a 65% help center engagement rate, alongside a 28% engagement rate among help center visitors, letting Borrowell scale support for over 3 million members without expanding the team.

Reported metrics

Resolution rate83%

Help center visitor engagement rate28%

Answer rate87%

Help center engagement rate65%

Show all 8 reported metrics

resolution rate83%

help center visitor engagement rate28%

answer rate87%

help center engagement rate65%

monthly ticket volume (baseline)more than 11,000 tickets per month

support team size (baseline)nine agents

peak monthly ticket volumebeyond 12,000 per month

members servedover 3 million members

Reported stack

Lyro AI AgentTidioZendesk

Source

https://www.tidio.com/blog/borrowell-case-study/

Read source ↗

Frequently asked questions

What did this team achieve with this AI workflow?

Lyro AI Agent achieved an 83% resolution rate, an 87% answer rate, and a 65% help center engagement rate, alongside a 28% engagement rate among help center visitors, letting Borrowell scale support for over 3 million…

What tools did this team use?

Lyro AI Agent, Tidio, Zendesk.

What results were reported?

Resolution rate: 83%; Help center visitor engagement rate: 28%; Answer rate: 87%; Help center engagement rate: 65% (source-reported, not independently verified).

How is this customer support AI workflow structured?

Lyro engages before ticket creation → Lyro answers from help center knowledge → Manual Q&A training for gaps → Member contact details collected → Unresolved chats become tickets → Full transcript attached → Ticket routed to support queue → Agent reviews with full context.