What is the difference between an LLM and a chatbot?
An LLM is the underlying model that turns input text into output text. A chatbot is a full product around it, with a chat interface, instructions, memory handling, safety checks, and often tools or company data.
Explain it simply
The LLM is the engine. The chatbot is the whole car: seats, steering, brakes and dashboard.
When you use a bank's support assistant, the LLM does the language work. Around it sit a system prompt that sets the tone, code that stores the conversation, a search over the bank's help pages, filters that block unsafe replies, and a handover to a human agent. Remove those and you have only a model that takes text and returns text.
This distinction matters in interviews because most real quality problems are fixed in the surrounding system, such as better retrieval or clearer instructions, and not by swapping the model.
Example
User message
-> input checks
-> add system prompt + chat history + retrieved help articles
-> LLM
-> output checks
-> reply shown in the chat windowWhat to say in the interview
"The LLM is one component. A chatbot adds the interface, prompt design, conversation memory, retrieval or tools, guardrails and logging. When I debug a poor answer I look at the whole pipeline, not only the model."