Generative AI, LLM, GPT, agentic AI - get used to the sound of these terms because you will hear them a lot through 2025. GenAI promises to be a gamechanger in the same lines as the wheel, the printing press, the steam engine, electricity, the personal computer, and the Internet.
While the rush to be known as an AI-enabled product or company is understandable, it is easy to get sucked into spending time, money, and resources on the wrong initiatives. I like to divide AI initiatives into two distinct types - external facing and internal facing.
External facing solutions are those with which your customers interact. ChatGPT and Gemini are examples of these. An AI-bot that interacts with customers and helps them complete a task is the holy grail. Like all fantastic-to-have features, it is also the most difficult to build properly. Before one embarks on this stone-and-thorn-littered road, we would be well served to go to years of Support conversations and identify the most common user actions. After identifying the list of actions, it will take more analysis before we begin to create a bot that is more than just a shiny new toy and brings actual value to the user.
Internal facing solutions are those that bring efficiency within the organization without actually interacting directly with the customer. For instance, a bot that constantly analyzes Support conversations and identifies opportunities for fixes, upgrades, or even new features. Such a solution may not be as glamorous as a bot that answers user queries, but it will improve the user experience by enabling better service.
While the temptation to reach for the shiny fruit on the tree is overwhelmingly tempting, it is good to remember that most often the ripe, ready-to-eat fruit are the ones lying at the base of the tree. Similarly, before putting the effort into building an AI-bot that interacts with customers, it is often a better idea to build the AI muscle by working on internal tools where failure is not as expensive. To make the deal sweeter, tools that increase efficiencies within the organization usually end up making more of a mark on the bottomline and their effects are easier to measure.
Generative AI is a tool; and like every tool, it is useful only when used appropriately, just like a hammer is great to beat a nail into a plank of wood but totally useless when it comes to making an omelette.
Bình luận