Sales assistance based on purchasing prediction

A method comprising:

• identifying, by one or more computer processors, one or more items searched for by a user on a client device by accessing the past browser history of the user;
• determining, by one or more computer processors, if a location of the client device is within a threshold distance of an item of the one or more searched items, wherein the location of the client device is identified by a Global Positioning System and the location of the item is stored in a database based on a planogram of the store;
• in response to determining that the client device is within the threshold distance, generating, by one or more computer processors, a communication that the item of the one or more searched items presents to the user of the client device, wherein the communication includes details about the item, personalized information about the user, and audio and visual components;
• sending, by one or more computer processors, the generated communication to the item to be displayed as text by the item;
• determining, by one or more computer processors, if a received response comprising a reply of the user to the generated communication indicates that a threshold probability that the user will purchase the item is exceeded, wherein the response is transformed to text via natural language processing, and wherein the probability is calculated via latent Dirichlet allocation;
• in response to determining that the received response exceeds the threshold probability, identifying, by one or more computer processors, a sales representative to assist the user;
• and
• sending, by one or more computer processors, an alert to the sales representative that the user intends to purchase the item, the alert identifying the item to be purchased.