The Hidden Cost of Broken Systems — And How AI Can Fix Them
The business case for AI is ever present.
Just this week:
1️⃣ I spent 90 minutes on a call to a major retailer working out a support issue with an appliance I had delivered.
2️⃣ My mother had a company working on a legal issue for her, but the case unintentionally dropped into a hole and I had to spend time getting that worked out for her.
3️⃣ We received a $18,000 medical bill in error because of an issue with insurance information being entered that I had to call the hospital to work out.
None of these issues were any specific person's fault. They were the result of a pattern of modern work where individuals are trying to keep too many plates spinning, and systems and processes are not built to rectify edge cases once they appear.
How would this look different with AI?
1️⃣ Provide in-call real-time agentic help to support agents so that issues outside of the norm can be quickly identified and routed for the right resolution, minimizing time to resolution and freeing up call center resources for more suitable cases.
2️⃣ Identifier outlier cases based on existing resolution data and initiate automated AI-based review to identify whether the blockers are ones that can be resolved in different ways. This gives your team an actionable to-do list rather than a sea of potential problems to fix.
3️⃣ Utilize past patterns to identify problems before they occur. If a family generally uses insurance, follow-up on why it's not entered on this case before the bill ever gets generated. Automate processes to get to the right solution rather than picking a route with a problem as the destination.
Now, obviously, none of these opportunities are just going to appear out of the box, the processes would need to be understood, and the solutions built. But another place AI can help is that by implementing tools like Microsoft 365 Copilot and getting back a couple of hours per week, you give your team the space and time to start looking at ways to tackle these big issues they know exist but rarely rise to the top of the urgency list.
By bringing together generalized AI tooling that helps knowledge workers to be more productive alongside targeted AI-first interventions that offer specific help to company- or industry-specific problems, we paint the full of picture of how this technology can help many businesses.
But we also need to remember, that we still must be careful about where and when to use AI. Across ChatGPT, Copilot, and Gemini if you ask for an image to go with this post of a "call center agent taking a challenging phone call" there is a universally logical but entirely mixed-up understanding of this request. Call center = headset. Phone call = handset. So, the image gets both. Understand these limitations.
Where are the obvious problems you run into in your work or when dealing with other businesses where an AI-based intervention is probably the best solution?
Posted on Linkedin on 08/01/2025 -> View Linkedin post here