{"id":21777,"date":"2024-04-29T15:43:18","date_gmt":"2024-04-29T15:43:18","guid":{"rendered":"https:\/\/www.insurance-canada.ca\/?p=80623"},"modified":"2024-04-29T15:43:18","modified_gmt":"2024-04-29T15:43:18","slug":"building-a-data-strategy-for-generative-ai","status":"publish","type":"post","link":"https:\/\/blog.lifeinsurance-orleans.ca\/index.php\/2024\/04\/29\/building-a-data-strategy-for-generative-ai\/","title":{"rendered":"Building A Data Strategy For Generative AI"},"content":{"rendered":"

By Manav Gupta, CTO & VP, IBM Canada \u2014<\/span><\/h4>\n

Artificial Intelligence is making headlines daily. It\u2019s dominating business conversations and shooting to the top of the agenda for many corporates, if not already there. According to IBM\u2019s Institute for Business Value, 60 per cent of C-suites will be piloting Gen AI solutions in their organizations by next year.<\/p>\n

Earlier this month I sat down with Bell\u2019s Director of IT Delivery, Neel Mehta, to discuss the integral connection between solid data foundations and an effective AI strategy. Neel gave a great analogy of this relationship \u2013 \u201cWe see data as the new oil, and AI as the refinery<\/em>.\u201d<\/p>\n

Bell\u2019s heritage is about connecting people to one another and to their devices. Today, they\u2019re using a combination of generative and predictive AI to serve their customers even further. For example, predicting extreme weather and environmental events which may cause outages, and getting their workforce ready for them.<\/p>\n

Neel shared an observation I\u2019m sure many other enterprises will relate to. That AI is galloping ahead at an accelerated rate, outpacing the rate that clean data can be fed in. That\u2019s where data strategy comes in, and a shift in focus from capturing data, to connecting data.<\/p>\n

The deployment of any new technology always comes hand in hand with implementation challenges. During our discussion we touched on what those have been for Bell, which I expect are shared with other organizations navigating their own AI journeys.<\/p>\n

    \n
  1. Data siloes<\/strong> \u2013 Like many enterprises, Bell has traditionally been organised vertically \u2013 a structure which doesn\u2019t lend itself to connecting data. The solution?  Building horizontal governance and data platforms so data can be shared between product lines and business units.<\/li>\n
  2. Data latency<\/strong> \u2013 AI is great for producing insights, but often the work is done through batch processes and the intel is sometimes late. Event driven architecture helps speed this up. Combining network events with customer journey events in real time enables Bell to target the right customer with the right device at the right time.<\/li>\n
  3. Data quality<\/strong> \u2013 Data governance is important, and the information feeding AI must be reliable. How can this be achieved? By cleansing the data and building the necessary framework to keep it that way when new data is curated.<\/li>\n<\/ol>\n

    Data quality is intrinsically linked to one of the hottest topics of AI conversation today \u2013 trust. It\u2019s important to keep a laser focus on keeping AI usage fair, unbiased and compliant. Customer transparency is just as important. They should know what you\u2019re doing with data and how long it\u2019s being kept.<\/p>\n

    We wrapped up our conversation on why it\u2019s important not to get carried away with the AI hype. It must solve real customer pain points, and outputs should always be linked to specific business goals. For organizations who are early in their AI journeys, or thinking of setting out, Neel shared this great advice:<\/p>\n