A Contact Centre Game Changer: QA + AI = Kwaai

Noori Allie

Artificial intelligence (AI) and data have the potential to radically transform contact centre quality assurance (QA). Where continuous improvement has traditionally been incremental, AI can drive progress in giant leaps.

In the world of customer loyalty programs, contact centres play a pivotal role in the customer experience. Each month, tens of thousands of calls come in, covering everything from points balances to password resets and product availability. 

Much like in sports, where the saying goes, "you’re only as good as your last match", customer satisfaction is often only as good as the last experience. Every call matters and has the power to strengthen or damage a customer’s relationship with a brand. This makes the role of QA essential - and with AI, the potential for improvement is boundless. 

The Current State of Contact Centre QA

Traditionally, quality assurers in contact centres:

• Listen to calls or recordings.
• Transcribe portions of those calls.
• Identify and address issues based on training and expertise.

However, there are significant limitations:

• A single QA professional can review approximately 250 calls per month, which is only 1% of calls in a contact centre handling 25,000 calls monthly.
• Call selection is random.
• QA primarily focuses on individual agent performance, making it difficult to derive broader insights at a team or macro level.

As Richard Cramer, Loyalty Director at Achievement Awards Group, notes, "We’re always looking for ways to ‘QA the QA.’ Most improvements are incremental. When something comes along that can 10X that improvement, you have to pay attention." 

The AI Advantage

What happens when AI and data analytics enter the picture? 

For starters, AI can transcribe all 25,000 calls almost instantly, expanding the sample size from 1% to 100%. Every call is analysed, removing the randomness from QA. 

AI can also be programmed to detect specific words, phrases, or sentiment cues, such as:

• Mentions of competitor brands—potentially signalling dissatisfaction.
• Signs of customer frustration, such as raised voices, fast speech, or swear words.
• Indicators of agent uncertainty, such as slow responses or low vocal energy.

Not only can AI detect red flags, but it can also highlight positive behaviours. Excellence that might otherwise go unnoticed can now be recognised and rewarded. 

At a macro level, AI can analyse aggregate data, identifying patterns across all calls. For example, if a large volume of customers are calling about password resets, that might indicate an issue with the user interface—one that, if fixed, could improve customer experience while reducing call volume. Likewise, frequent inquiries about a particular product’s availability could signal an opportunity for strategic partnerships or improved inventory management. 

Another AI advantage: It never forgets. Unlike human employees who require training and re-training, AI continuously learns from past data. It can track shifts in customer behaviour over time—such as changes in call frequency or peak call times—allowing businesses to proactively adapt.