A AI agent in e-commerce It’s an artificial intelligence system that can do that Independently make decisions and take actions to achieve a business goalB. sell more, serve better or reduce friction by using real-time data. Agents will now be commonplace. This happens from two points of view: On the one hand, sellers will be able to make better use of communication and support for their customers. On the other hand, it will be the consumers themselves who will “delegate” the perception of the environment to these “robots”, decide and act.
In this sense, more and more solutions are already emerging. Tiendanube has launched the first global platform to enable automated sales via WhatsApp through a strategic alliance with Meta. This functionality is integrated into Chat CloudTiendanube’s conversational commerce tool, which uses advanced artificial intelligence to optimize customer service. “You have to degrade the entire context,” explains Martín Palombo, co-founder of the platform. “The agent responds and internally determines how the query is used,” he adds. The idea is that With the information, the conversation develops “naturally” and the process can be completed by the time of payment.
The agent is able to do this Interpret text, images and audio in customer requests. View the store catalog and respond in real time with available information. And select items, assemble the shopping cart and prepare it for payment. They estimate that 65 percent can be answered by AI without human intervention. And the cost of the service is calculated based on the number of conversations.
Other companies are already implementing these solutions, which are very different from traditional bots. In the case of Botmakeran Argentine conversation solutions company with the largest distribution in Latin America and pioneer in the integration of the ChatGPT platform, offering the opportunity to companies of different sizes and industries Deploy artificial intelligence agents in channels such as WhatsApp, websites and emailamong others, both through text and voice, integrating artificial intelligence throughout the entire process of a potential service or transaction.
The idea is to provide automated assistance tailored to each customer’s tone, language and urgency, for example through generative audios on the platform. Or improve user engagement through more agile, fluid, and human experiences. You have already had some experiences, e.g. B. with Megatone tool housewhich implemented massive sending of WhatsApp notifications with discounts and promotions that increased traffic in physical stores. Or with Iplan, which has developed an assistant for contract processing directly via WhatsApp. The system validates the coverage area, accepts the user’s data and automatically concludes the contract. Every fifth monthly turnover is made with this assistant.
On the other hand, the idea is to write a prompt (Guidance) by an AI (it can be ChatGPT, Gemini) and that an agent is responsible for conducting the search, integrating options and providing purchasing solutions. Gabriela Renaudo, Country Manager Argentina and Southern Cone, spoke at the last Fintech Forum visa expects what will be Agent trading. “Today we use AI to improve the search for a product, but when it comes to paying we go somewhere else. Here we delegate to an agent who buys and pays for you. Purchase concert tickets, tickets and accommodation and products only with an instruction.
In the same sense, Mastercard also announced its own “Agent Pay.” that integrates payment experiences into the personalized recommendations and insights offered on conversational platforms and enables digital agents to process, manage and optimize shopping experiences on behalf of users. In the background, there are security measures that accompany these processes: dynamic credentials, fraud prevention and tokenization.
But far from being a future without controversy, some problems have arisen and we will have to see how much they slow down this possible future. A very resonant case is that mentioned by Alan Daitch. Amazon sued Perplexity, an artificial intelligence startup, to prevent its bots from crawling its website and help users select and purchase products. The conflict goes beyond a specific lawsuit: it reveals a fundamental struggle between the control over information and advertising that the major search engines and platforms want to retain, and AI’s promise to give decision-making power back to users.
Perplexity has developed an assistant capable of browsing Amazon, comparing options and even making purchases on behalf of the user.. Amazon requires these bots to identify themselves in order to block them. Which, of course, would be at odds with that promise: that of agents being able to make decisions on other platforms that don’t authorize those interactions.
About The Author
