Introduction to Vector Databases
Prompt Engineering is the art of designing effective inputs (prompts) to get the most accurate, relevant, and helpful responses from a Large Language Model (LLM). Since LLMs respond based on the text they receive, carefully crafting your prompt directly impacts the quality of the output.
What is a vector embedding?
A vector embedding is a way of converting text, images, or other types of data into a numerical representation—usually a list of floating-point numbers (vectors).
These embeddings allow AI models to understand meaning, context, and relationships between different pieces of text.
Example: The sentences
• “How’s the weather today?”
• “What’s the forecast for today?”
might have very similar embeddings, even if the wording is different. This makes it easier for AI to recognize that both are asking the same question.
Why use vector databases for AI?
Traditional databases are good at exact matches — for example, finding a product by its ID. But AI applications need more than that: they need to search by meaning, not just exact keywords.
Vector databases enable:
• Semantic Search — Find results based on meaning, not just keywords.
• Contextual Chatbots — Retrieve relevant past information to provide smarter answers.
• Question-Answering Systems — Match user questions with the right information, even if phrased differently.
Use cases include:
• Chatbots that search across documents
• Product recommendation engines
• AI assistants for technical support or FAQs
