Bitrix24 shows how to add four automation robots to its business portal

Bitrix24 has released the next installment of its hands-on development series for the business portal, this time focusing on automation. In the new article, the team shows how to add four robots to the system that bring data into a uniform format and collect key metrics on tasks and deals.

Which robots were added

This is not an abstract demo but a practical set of automations for CRM and workflows. The authors take the front-end app built in the previous part and extend it with robots that can be embedded directly into portal scenarios. The logic is simple: the cleaner the source data and the fewer manual actions employees perform, the easier it is to build reports and launch the next funnel stages without constant manual checking.

• Phone number correction — the robot converts contacts to a single format so they can be used correctly in CRM and downstream scenarios.
• Name normalization — the automation removes inconsistencies in spelling and helps keep customer and contact records cleaner.
• Task data collection — the robot pulls the metrics needed for work monitoring and internal analytics.
• Deal amount calculation — the scenario gathers financial data to assess the funnel and managers' workload faster.

What the project is built on

The project uses a prepared AI starter — a project template designed for development alongside AI agents. It already includes structure, instructions, and boilerplate that help you get started on the task faster without spending time on basic setup.

In effect, Bitrix24 offers not just a set of automation ideas but a starting environment in which to reproduce them, adapt them to your own portal, and test them on a real scenario.

This approach matters for teams just starting to build internal extensions for Bitrix24. Instead of a blank page, the developer gets a clear path: take the template, spin up the project, add the needed robots, check the data processing, and then bind the automation to specific CRM entities. One more plus: the article promises to show not only the final result but every change step by step, which lowers the barrier to entry for anyone who has not worked with such scenarios before.

Where difficulties arise

"Today we will show how to build automation and add 4 robots: phone

number correction, name normalization, and data collection for tasks and deal amounts."

The description itself shows where problems most often appear. Any automation depends on the quality of the source data: phones are stored in different formats, names are entered with errors, and the needed metrics sit across several entities at once. So even small robots need careful validation, clear normalization rules, and a check that the data remains usable for CRM logic and reporting after processing. This is usually where first implementation attempts break.

The authors specifically stress that the material covers possible problems during the work and ways to solve them. That is an important signal: the article targets not just a "happy path" demo but real development, where you have to fix configuration, catch integration errors, and adapt scenarios to the structure of a specific portal. For practicing teams, this is the value: you can see typical failures right away and understand how to avoid getting stuck at the implementation stage.

Another practical bonus: the project repository is available alongside the walkthrough. That means the material can serve not only as reading but as a basis for your own prototype: compare file structure, repeat the changes, and move individual blocks of logic into your corporate scenario without lengthy preparation. For teams, it is a convenient way to move quickly from theory to a working pilot inside their own portal.

What this means

Bitrix24 is gradually turning AI from a general conversation into a set of concrete business scenarios. For companies that live inside CRM, this is a useful signal: automation starts not with "smart agents in general" but with small robots that clean data, calculate metrics, and take manual routine off standard operations.