What was done
TELUS moved its contact-centre analytics and conversational agents onto Google Cloud. Gemini-based agents now handle a share of customer calls end to end, while the platform reviews every recorded interaction instead of a small sample, so the team can spot the root causes behind calls. Frontline staff also lean on it for ticket creation, call summaries and lookups.
Outcome
Google Cloud reports that about 30% of customer calls are now resolved proactively by AI agents, issues get resolved 87% faster, and TELUS saves $53.9M a year in operating costs. Call analysis grew from roughly 200K to more than 22M calls a year. These are vendor-reported figures.
Evidence: Vendor case study. Figures are as reported by the source and have not been independently audited by AgentsNorth.
Prompt Reconstructed by AgentsNorth
The source didn't publish its prompt. This is our reconstruction of a starting prompt for a similar workflow, not the company's actual prompt.
You are a customer care agent for a Canadian telecom provider. Greet the caller, confirm their identity using the account verification steps, and identify the reason for the call. For billing, outage or plan questions, look up the account and resolve the issue directly, explaining each step in plain language. If the issue needs a technician, a retention offer or a complaint escalation, summarize the call (reason, account, steps taken) and transfer to a human with that summary. Never promise credits you can't apply.