Cómo escribir una carta de recomendación con IA

A strong recommendation letter does three things a boilerplate letter cannot. It establishes the writer's basis for judgment, gives specific evidence and makes an honest comparison. The standard structure opens with who you are, how you know the candidate, for how long and in what role. The body carries two or three anecdotes, each showing a specific action and its result, chosen to match what the program or employer says it values. The close gives a clear endorsement and your contact details. AI helps most with gathering and shaping. Before drafting, collect the candidate's resume and the job posting or program description. Ideally, also ask the candidate for a short 'brag sheet' listing their projects and what they want emphasized. Then ask the AI to interview you about moments you remember. A line like 'she stayed late to fix the data pipeline before the grant deadline, then wrote documentation so the next analyst could run it' is far more persuasive than 'hardworking and dedicated.' The biggest misconception is that AI can make a letter stronger than your actual knowledge. It can make weak material sound polished, but experienced readers recognize adjective-heavy praise without evidence. Admissions committees, for example, read hundreds of letters. Comparative statements, like ranking a student among the best you have taught, carry weight only if they are true, and you are signing your name to them. Two more cautions. First, letters describe another person, so educators should think carefully before pasting student records into consumer tools. In the US, student education records are protected under FERPA. Second, watch for gendered language. A well-known 2003 study by Trix and Psenka looked at recommendation letters for medical faculty. It found that letters for women were shorter and more often emphasized effort than accomplishment. Asking the AI whether your letter describes achievements or only diligence is a quick safeguard.