Glossary
Tool use. Tool calling.
The capability of a language model to invoke external functions, APIs, or systems as part of generating an answer. Instead of just producing text, the model can search a database, call a service, run a calculation, or update a system, and use the result in its response.
A model without tools can only describe what should happen. A model with tools can make it happen. Tool use is what turns a language model from a writer into a doer.
What this looks like in practice
Why tools matter for accuracy.
A model with access to a calculator doesn't have to guess at arithmetic. A model with access to your customer database doesn't have to invent customer details. A model with access to your knowledge base doesn't have to hallucinate from training data. Tool use is how you get from "what the model thinks is true" to "what is actually true."
The hard part: knowing when to call which tool.
Giving a model three tools and asking it to pick the right one for each step is a small problem. Giving it thirty and asking the same is a hard problem. The model has to understand each tool's purpose, when it applies, what parameters to pass, and what to do with the result. Most off-the-shelf agent frameworks struggle past five or six tools.
The trap: tools that succeed but produce nonsense.
A tool call can return a successful response that's still wrong. A SQL query runs, returns rows, but the rows don't actually answer the user's question. The model dutifully reports the result as the answer. Tool use without verification is just hallucination with extra steps.
Tools turn description into action.
Simply Asking gives Lumen the tools it needs to act on your knowledge: search your library, query your data, update your records, send notifications. Every tool call grounded in your business context, every result verified before it's used in an answer.
Bring your knowledge.Simply ask.
CASA Certified · SOC 2 Infrastructure · GDPR-Aligned · Enterprise Ready