Definition
Large Language Model (LLM)
A large language model is an AI system trained on vast amounts of text to predict and generate language, enabling it to answer questions, summarize, write, and reason over natural language. LLMs power modern AI assistants and agents. They are probabilistic next-token predictors, which makes them remarkably capable but also prone to confident errors.
Key takeaways
- A large language model is trained on vast text to predict and generate language — answering, summarizing, writing, and reasoning over natural language.
- It works one token at a time within a bounded context window, so supplying the right context is half the engineering.
- LLMs are probabilistic, not deductive: they can be fluent and wrong (hallucinate), which grounding in retrieved data helps prevent.
- In real products they're wrapped with tools, memory, and retrieval — turning a text predictor into an agent that can take action.
At its core, an LLM does one thing: given some text, it predicts what comes next, one token at a time. Trained on enormous and diverse text corpora, it learns the statistical structure of language so deeply that this simple objective yields surprisingly general capabilities — translation, summarization, coding, step-by-step reasoning — none of which were programmed in explicitly. The model is a compressed, queryable representation of patterns in its training data.
Two properties shape how LLMs are used in products. First, they work within a context window — a bounded amount of text they can consider at once — so feeding them the right context is half the engineering. Second, they are probabilistic, not deductive: they generate plausible continuations, which means they can be fluent and wrong, producing confident fabrications known as hallucinations. Grounding their answers in retrieved, authoritative context is the main defense.
In applications, LLMs rarely act alone. They are wrapped with tools they can call, memory of prior steps, and retrieval that pulls relevant data into context, turning a text predictor into an agent that can take real action. The model supplies reasoning and language; the surrounding system supplies facts, capabilities, and guardrails.
Planoda uses LLMs through a model gateway with retrieval grounded in your workspace and a propose/approve guardrail on consequential actions, so AI features draw on real, tenant-scoped data and act as accountable teammates rather than unsupervised text generators.
Related terms
- AI AgentAn AI agent is a software system that uses a large language model to pursue a goal across multiple steps — reading context, choosing tools, and taking actions — rather than answering a single prompt. In a work platform, agents triage issues, draft updates, and execute multi-step tasks as autonomous teammates, bounded by the permissions and approvals their operators set.
- Retrieval-Augmented Generation (RAG)Retrieval-augmented generation (RAG) grounds a language model's answers in your own data by retrieving the most relevant documents at query time and feeding them into the model's context before it responds. Instead of relying solely on what the model memorized in training, RAG lets it answer from current, authoritative, private sources — sharply reducing hallucination.
- Context WindowA context window is the maximum amount of text — measured in tokens — that a language model can consider at once, spanning its input and its output. Everything the model 'sees' for a given request must fit inside it: the prompt, retrieved data, conversation history, and the answer. When content exceeds the window, the oldest or least relevant parts must be dropped.
- Semantic SearchSemantic search finds results by meaning rather than exact keywords, using vector embeddings that place similar concepts near each other in mathematical space. A query for 'login broken' can surface an issue titled 'users can't authenticate' even with no shared words. It powers more relevant search and is the retrieval layer behind many AI features.