> For the complete documentation index, see [llms.txt](https://help.whaly.io/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://help.whaly.io/core-concepts/data-modeling/maintaining-data-models.md).

# Maintaining data models

## Iterating on your models

As building models can be an overwhelming  endeavour, it is important to take things one step at a time. Iterating on your models will help you mitigate some of this complexity by slicing a big work package into smaller and simpler tasks

#### Start small 

Start by creating the minimum models you will need. Select the bare minimum number of columns you need and start from there.

#### Add what's needed

When you need to add a new information to a model, you should always ask yourself the following question : Should I extend my existing model or create a new one. If you sense that adding a new column will change the purpose of your current models, you should probably create a new one. If not, you should update your model.

#### **Carefully remove what's not needed anymore**

Removing columns may be a tricky operation, as your model columns might be used in other part of the BI (explorations, relationships, drills and even other models). If you need to replace a column, we advise you start by creating a new column, update your model and it's dependencies and then remove the column. This should avoid some downtimes and mitigate risks to break downstream items. 

#### Repeat

Iteration is a never ending process, so don't be afraid to iterate every time it's needed

## To go further 

Building models in Whaly is a great way to quickly deliverer value and iterate on feedback. When some of your Whaly models are widely adopted and become a central part of your BI analysis, you should think about moving them to a more stable and testable environment, such as [dbt](https://www.getdbt.com/).

Some core functions of dbt are building, testing and documenting SQL models, as well as providing a complete version control system.

Whaly offers a direct integration with dbt cloud: you will be able to automatically import your dbt models in your Whaly environment in order to use them in your explorations and your charts.