How to implement AI in the contact center: Nine lessons from CX Shift
Key takeaways
- AI is most effective when it removes agent struggle rather than trying to replace the agents themselves.
- Digital tools should handle simple transactions so humans can focus on high-stakes emotional relationships.
- New AI simulations act as "flight simulators" to prepare agents for high-pressure calls before they go live.
- Agent turnover is an expensive but solvable problem that costs companies roughly $20,000 per seat.
- Coaching needs to shift from quarterly performance reviews to real-time, daily AI mentorship.
- Hiring models must evolve to prioritize empathy and critical thinking over basic script reading.
At Shift: The human + AI CX summit, customer experience leaders with decades of contact center experience discussed how humans and AI are working together today. As apps, chatbots and self‑service tools handle routine tasks like order status and password resets, the interactions reaching human agents are fewer, but much more complex and emotional.
Shift focused on what is actually working right now and revealed clear lessons for using AI to support agents in winning the moments that matter.
Here are nine of the most important takeaways CX leaders can apply today.
1. AI should eliminate struggle, not eliminate people
The strongest message of the summit was that AI succeeds when it removes friction for agents instead of replacing them. Bret Kinsella, SVP/GM of Fuel iX™ at TELUS Digital, noted that organizations should choose to use AI to "eliminate the struggle" rather than the person.
Leaders shared that agent burnout is driven by cognitive overload — not call volume alone. AI that handles search, summarization and guidance directly reduces that load.
Start AI adoption with agent pain points, not automation goals.
2. Digital handles transactions. Humans handle relationships
As self‑service improves, the calls that reach live agents are far more complex. While ~60% of customers prefer digital for simple needs, 100% want a human when loyalty is on the line.
Design AI to triage, prepare and support agents in high‑stakes conversations.
3. Agent experience and customer experience are inseparable
Better CX does not come from pushing agents harder. It comes from preparing them better. Confident, well‑trained agents naturally improve AHT, CSAT and first‑call resolution as a byproduct of better conversations.
4. Simulation training dramatically improves time to proficiency
Traditional onboarding often fails to prepare agents for real customer pressure. Fuel iX Agent Trainer, for example, uses simulated voice and chat to accelerate proficiency by up to 50%.
The summit highlighted the “90‑day cliff”, where many agents churn during training or shortly after going live.
Use AI simulations that mirror real, high‑stakes scenarios before agents take live calls.
5. Coaching must move from quarterly policing to daily support
Retrospective QA scorecards slow learning and damage morale. The industry is moving from a user experience that monitors to one that mentors.
Teams using real‑time feedback and micro‑learning see faster skill improvement than teams relying on quarterly reviews.
6. Agent attrition is a solvable business problem
Turnover is not inevitable — it is often the result of poor preparation and constant stress.
Some industry data indicate that the cost of filling an agent's seat is $10,000 to $20,000 per agent. For a 100‑seat contact center, a 45% attrition rate can cost millions annually.
Use AI to support agents before burnout sets in, not after they fail.
7. The agent role is evolving faster than hiring models
Today’s agents need judgment, empathy and critical thinking. There is a clear shift away from basic digital skills toward contextual decision‑making and emotional intelligence.
8. AI is never “set and forget” in the contact center
"Everything needs curation," says Erin Walker, Global VP at TELUS Digital. Products, policies and customer expectations evolve faster than static workflows.
Assign clear ownership for continuous AI tuning and improvement.
9. The future of CX is human
AI is accelerating change, but it is not changing what customers value most. The most successful organizations use AI to elevate human connection, not distance customers from it.
Make human + AI collaboration the north star of your CX strategy.
Watch all nine info-packed sessions led by the best CX experts in the business.
