[{"data":1,"prerenderedAt":84},["ShallowReactive",2],{"glossary-en-prompt-engineering":3},{"id":4,"title":5,"body":6,"description":71,"extension":72,"meta":73,"navigation":79,"path":80,"seo":81,"stem":82,"__hash__":83},"en_glossary/en/glossary/prompt-engineering.md","What is Prompt Engineering?",{"type":7,"value":8,"toc":62},"minimark",[9,14,21,26,29,51,55],[10,11,13],"h2",{"id":12},"prompt-engineering","Prompt Engineering",[15,16,17,20],"p",{},[18,19,13],"strong",{}," is the art and science of designing and optimizing inputs (prompts) to get the best result from a model.",[22,23,25],"h3",{"id":24},"key-concepts","Key Concepts",[15,27,28],{},"It involves more than just asking a question. It includes:",[30,31,32,39,45],"ul",{},[33,34,35,38],"li",{},[18,36,37],{},"Context Setting",": \"You are an expert industrial automation engineer.\"",[33,40,41,44],{},[18,42,43],{},"Constraint Definition",": \"Answer in JSON format only.\"",[33,46,47,50],{},[18,48,49],{},"Few-Shot Learning",": Providing a few examples of the desired input-output format within the prompt.",[22,52,54],{"id":53},"application","Application",[15,56,57,58,61],{},"In automated systems, prompts are often pre-engineered \"under the hood\". For example, a future feature in ",[18,59,60],{},"ZMA software"," might use a hidden, carefully crafted prompt to ask an LLM to \"Analyze these vibration statistics and summarize the health status in 3 bullet points\" ensuring the output is always consistent for the end-user.",{"title":63,"searchDepth":64,"depth":64,"links":65},"",2,[66],{"id":12,"depth":64,"text":13,"children":67},[68,70],{"id":24,"depth":69,"text":25},3,{"id":53,"depth":69,"text":54},"The art of designing inputs to get the best possible output from an Artificial Intelligence model.","md",{"tags":74},[13,75,76,77,78],"AI","LLM","Optimization","Interaction",true,"/en/glossary/prompt-engineering",{"title":5,"description":71},"en/glossary/prompt-engineering","1l0Y0KJR2poBw5tM2h1NfXRbiKK2zin-5HpZqQV8G1I",1778229654972]