Glossary
What is a large language model (LLM)?
A large language model, or LLM, is an AI system trained on very large amounts of text to predict the next words in a sequence. That simple task lets it write, summarise, translate and answer questions in fluent language. Claude, GPT and Gemini are examples. In sales tools, an LLM drafts messages and reads replies.
Published by Sellinger AI. Numbers and settings checked on 24 October 2026 against the fact register. Report an error.
What does an LLM do well in sales?
- Drafting a first message from facts about a person and a company.
- Summarising a long thread or a company website.
- Reading a reply and sorting it: interested, not now, wrong person, opt out.
- Writing in the prospect's language.
- Turning rough notes into a clear follow up.
Where do LLMs fail?
They can state things that sound right and are not, which is called hallucination. They know only what they were trained on and what you give them, so a model without your price list will guess a price. And they write fluently even when the input is thin, which can hide a poor list or an unclear offer. Good tools reduce these risks with retrieval from your own documents, rules and human review.
What should you check in a tool built on an LLM?
| Question | Why it matters |
|---|---|
| Where do the facts in a message come from? | Messages should use your knowledge base and the prospect's public data, not guesses |
| What happens when the model does not know? | It should hand over, not invent |
| Can a person approve messages? | Review catches what rules miss |
| Can you read every message sent? | You are responsible for what goes out in your name |
How Sellinger handles it
The agent researches each prospect, writes the first message and every follow up, answers the replies and books the meeting on your calendar. It answers from your knowledge base, your persona and your rules. A campaign can run in approval mode, where every message waits for a person to approve it, or send automatically. A person can take over any conversation, and the agent stops on its own when it cannot answer, for example a price it has no answer for.
You can also work with Sellinger from Claude or ChatGPT through the MCP connector.
Common questions
Is an LLM the same as an AI agent?
Which LLM is best for sales?
Do LLMs learn from my messages?
Sources and method
General AI practice. Sellinger behaviour from the fact register, checked 24 October 2026.
Related: AI hallucination, Retrieval augmented generation, AI sales agent, ChatGPT or Claude to run outreach.
