---
title: "function calling | Skills We Assign For | Azendo"
description: "Function calling for model-driven actions — schema design, validation and the Azendo roles assigned for AI work."
url: "https://azendo.co/skills/function-calling/"
---

[Skills](https://azendo.co/skills/) AI and machine learning 

# function calling.

Function calling lets a language model return a structured request to run a named function with arguments, rather than free text. It is the mechanism that turns a model from something that writes into something that acts.

## Where function calling fits on a long engagement.

The model does not execute anything. It proposes a call, your code decides whether to run it, and that boundary is the whole safety design. Treating a proposed call as an instruction rather than a request is how these systems cause damage.

Reliability comes from narrow, well-described functions with validated arguments. A broad function with a vague description gets called at the wrong time with plausible-looking nonsense, and anything irreversible needs a human approval step regardless of how well the model performs in testing.

## What an assigned team does with function calling.

The boundary between what a model proposes and what your code executes is a design decision that gets revisited every time a new capability is added. Each addition widens what the system can do and what it can do wrong, and that trade needs an owner.

Where the goal is internal process rather than a customer-facing feature, the whole assignment sits under [AI automation engineering](https://azendo.co/services/ai-automation-engineering/), with approval steps designed in from the start.

## What we use function calling for.

* Assistants that do work A model that can look something up or file something, rather than only describing how to.
* Structured output from unstructured input Extracting fields from documents or messages into a validated schema rather than parsing prose.
* Guarding irreversible actions Approval steps built into workflows where an error would be expensive to undo.

## How function calling capacity is assigned.

This work is assigned under AI automation engineering where the goal is internal process, and under AI engineering where it is a product feature. What that figure covers is set out in [how the monthly fee is built](https://azendo.co/pricing/).

## Roles we assign function calling for

* [AI Automation Engineer AI automation engineering](https://azendo.co/services/ai-automation-engineering/ai-automation-engineer/)
* [AI Engineer AI engineering](https://azendo.co/services/ai-engineering/ai-engineer/)

## Service lines it sits in

* [AI automation engineering](https://azendo.co/services/ai-automation-engineering/)
* [AI engineering](https://azendo.co/services/ai-engineering/)

Capacity is agreed as a committed monthly capacity across a discipline, not per skill.

## Related in ai and machine learning

* [PyTorch — skill we assign for](https://azendo.co/skills/pytorch/)
* [TensorFlow — skill we assign for](https://azendo.co/skills/tensorflow/)
* [scikit-learn — skill we assign for](https://azendo.co/skills/scikit-learn/)
* [Pandas — skill we assign for](https://azendo.co/skills/pandas/)
* [NumPy — skill we assign for](https://azendo.co/skills/numpy/)
* [embeddings — skill we assign for](https://azendo.co/skills/embeddings/)
* [vLLM — skill we assign for](https://azendo.co/skills/vllm/)
* [MLflow — skill we assign for](https://azendo.co/skills/mlflow/)

## Tell us what your roadmap needs function calling for.

A service delivery manager replies with the disciplines we would assign, the monthly capacity and what the first month looks like.

[All skills we assign for](https://azendo.co/skills/)