Optimize a Workflow

As you deploy automation across your work or business, optimization is an important goal for new and existing Workflows and we’ve made it simple to keep things running fast, dependably, and at scale.

Introduction

Catalytic is purpose built for people-friendly automation and our app builder is low-code to make automation simple. Because we keep things simple, your optimization needs will vary depending on each Workflow’s structure and complexity.

When we talk about optimization, we’re referring to simple optimization principles that help decrease Workflow duration, or how long each Workflow takes to complete. Generally, all Workflow processes will benefit from applying optimization principles, but particularly automation that is more complex or robust.

If a Workflow that started out simple grows in complexity, continue returning to these optimization principles to keep work running smoothly; or, keep these ideas in mind during your next build.

Benefits of Optimization

Duration is an important automation measurement and a great way to see the benefits of optimization. When you optimize your Workflows, they run faster.

With Catalytic, you can improve duration by optimizing how many actions, tasks, and subprocesses are included in a Workflow, and building with system limits in mind.

If you have a process where you can eliminate 50% of the actions, you’re likely to see a significant improvement in Workflow duration. If you’re eliminating only 5% of the actions in a process, there will be an improvement, but not as significant of one.

There are 3 optimization principles: reduce extraneous work, combine similar actions, and build with system limits. If you can apply all 3 to a single Workflow, improvements could be remarkable!

Reduce Extraneous Work

Triggering subprocesses to run from each row of a spreadsheet or data table using the Tables: Start Workflow for Each Row action is a commonly used building technique and a great way to get a lot of automation moving quickly.

Since the Tables: Start Workflow for Each Row action starts a Workflow for each row of a data table, extra work can sneak in. You can reduce this extraneous work by filtering the table to remove unnecessary rows first.

screen readers look here

Use actions like Tables: Apply Filters or Tables: Get unique rows first to filter and remove rows, thereby reducing the number of triggered subprocesses. This way, as your spreadsheet or data table grows, the process will run as fast as possible since only the essential subprocesses are triggered.

Combine Similar Actions

When a Workflow is first designed, the actions or flow may seem to be most efficient way of automating work, but as we add new actions and ways of automating, look back through a Workflow again to find opportunities to optimize.

For example, if text or data is being cleaned up to be parsed or input into another system, steps like removing characters, bullet points, or spaces may be added back to back as separate actions or appear multiple times throughout a process inadvertently.

screen readers look here

Combining these into one action, like with the field formulas action, or by sanitizing the data earlier in the process so it’s only performed once, reduces the total number of actions.

Build with System Limits

There are various system limits enforced by Catalytic that ensure a high-quality experience for you and your team. These limits help prevent interruptions to your processes and automation.

Most limits are managed behind the scene, like if too many tasks are started at once, the work will be smartly queued and completed as soon as possible. If you have a Workflow that consistently exceeds system limits the process runtime will be impacted.

One way to build with application limits and improve Workflow duration is best illustrated by the current limit on the max number of tasks that can be started at once:

  • There is a current system limit of 10 tasks started per second per team. If you have a Workflow that starts 15 tasks per second per team, the first 10 tasks are started immediately, and the remaining 5 will be queued to start right after. This process will then take longer than if the process was set up with the system limit in mind.

In the Workflow system limits article, specific Workflow system limits are explained for features like data tables, excel files, or tasks and actions.