Function calling / tool use
Function calling enables AI models to invoke external tools, APIs, and functions as part of their reasoning, allowing them to not only generate text but.
Detailed explanation
An LLM without tools is like a brilliant employee who isn't allowed to call, email, or use their computer. Function calling breaks this limitation by giving the model the ability to call external systems: look something up in a database, send an email, update a CRM record, retrieve the weather, calculate a price. In the context of AI sales agents, function calling is the mechanism that allows AI-agent to actually perform actions instead of just generating text.
Synonyms
Examples
AI-agent reasons: 'This prospect has shown interest but hasn't scheduled a meeting yet. I need to send a follow-up.' AI-agent sequentially calls: get_prospect_data → get_email_history → generate_followup_email → send_email → update_crm_status. All in one reasoning chain, fully autonomously.
A function call for lead enrichment: AI-agent searches a prospect via LinkedIn scraper, enriches data with company information, checks technology usage, and generates a hyper-personalized email based on all enriched data.
When to use this?
Function calling is essential for AI agents that need to not only think but also act, it's the difference between an AI that gives advice and an AI that actually does the work.
Your Company OS approach
With function calling, an AI colleague in Your Company OS can perform actions in your connected systems, such as a CRM update or a calendar entry. What it may do is defined in advance; sensitive actions remain subject to approval.
Related terms
Want to know more about Function calling / tool use?
Want to know how to put Function calling / tool use to work in your organisation? Get in touch with Your Company OS.