Why Data Quality Should Be Your AI Mantra

Data quality is the bane of many AI implementations. It’s a bit similar to cooking—if your pulses are stale, your daily comforting bowl of daal will be anything but that. Getting data right is key to gaining the right insights from AI. This was brought home to me though a recent experience with Airtel, a company armed with cutting-edge tech and impressive customer-experience capabilities. We were shifting home and so I put in a shifting request. Now, we have two Airtel broadband connections (one in a different city), so I was careful to mention the right landline number and WiFi ID that Airtel uses to identify connections. It was a simple shift within the same apartment block and when the Airtel technicians came two days later, all seemed fine till they couldn’t activate the connection on their app. They left saying they’d be back in 30 minutes. Hours passed and when I called Airtel to figure out what had happened, they seemed to have no clue. I was just promised they would set things right, which as with most promises made by agents at contact centres, turned out to be meaningless. Almost three days later, after multiple escalations and a slew of broken promises I figured out what had happened. Despite me being clear on the connection to be shifted, the Airtel customer care executive had inadvertently raised a shifting request for the other number. When I realised this, the first thought I had was that for all its tech, Airtel really needs to sort its human intelligence as opposed to only touting their spam-fighting AI, because over this entire period they could not figure out this simple error by a human. I’m on Airtel Black, which supposedly delivers premium customer service and the service on display was akin to what I experienced with MTNL when growing up—when there was no competition and 10 years to get a landline was normal. I know Airtel’s service quality is far, far superior to what MTNL delivered back then, but that’s how it feels like when you’re stuck with no connectivity and the company tasked with connecting you is utterly clueless about what has gone wrong. But enough of my woes and back to AI. To me this entire saga underlined how one piece of bad data (in this case, the wrong landline number or Wi-Fi ID chosen by a likely inattentive agent) can result in chaos and the opposite of the experience that a telco would seek to deliver to ensure customer loyalty in a very competitive market. Data quality should be the mantra that CXOs deploying AI solutions should live by when it comes to these implementations. And when that basic building block is done right, organisations can focus on leveraging AI as a growth catalyst as opposed to just a tool for productivity. One of our stories on AI this week talks about how some companies who are AI leaders thanks to their sharp and disciplined focus on getting AI right are applying AI to increasingly sophisticated areas, such as decision-making and workflow orchestration. And by doing so they are unlocking greater value than some of their peers. .Are You Trying Too Hard To Change For AI?.Here are some of our other AI-related reads from the week: API Holdings Powers Digital Health Technology Stack With IBM Instana AI Solution. Still 'Google It': AI Chatbots Like ChatGPT, Gemini Made Up Just 2.9% Of Search Traffic In 2024 Alarming Lack Of Security Readiness: Mere 4% Enterprises 'Mature' Enough To Counter Cyber Threats Google Cloud, Philips To Leverage AI To Drive Consumer Marketing Innovation Agentic AI, IoT Security, Humanoid Robots Among Forrester's Top Emerging Technologies For 2025 Truecaller Message IDs: AI-Powered Feature To Cut Through Spam, Identify Critical SMSes AI To Help Prevent iPhone Battery Drain: Apple May Launch Battery-Saving Feature With iOS 19 Google's Search Dominance Faces A Real Test In The Age Of Generative AI: Report Apple Announces AI-Powered Accessibility Features: Nutrition Labels, Mac Magnifier App, Braille Access Smarter Wearables: Google Gemini AI To Debut On Samsung Galaxy Watch; Buds Experience To Improve Moto Razr 60 Ultra With Advance AI Launched — Check Features, Price In India .Till next week, Ivor Soans .Subscribe to our newsletter here, to get this in your inbox..Could AI Entice You Into An Unhealthy Relationship?

Why Data Quality Should Be Your AI Mantra

Data quality should be the mantra that CXOs deploying AI solutions should live by when it comes to these implementations.

NDTV Profit

NDTV Profit

Follow Us

DOWNLOAD THE APP