How to automate processes?

Overview

01–02: Select and map processes
03–04: Analyse, redesign, and design
05–06: Implement and deploy
07: Monitor and optimise

1. Identify and select processes

Analyse existing processes

Collect information about the processes currently used in your company.

Select processes for automation

Decide which processes are suitable for automation using criteria agreed in advance: cost, time, repeatability, benefits, and risks.

2. Document and map processes

Process mapping

Create visual process maps, for example using BPMN diagrams, showing individual steps, decision points, and responsibilities.

Process documentation

Describe the steps, inputs, outputs, and resources required in detail.

3. Analyse and redesign processes

Process analysis

Identify weaknesses, bottlenecks, and opportunities for improvement. Define how to assess the efficiency and effectiveness of existing processes.

Process redesign

Develop improved processes that address the identified shortcomings and prepare them for automation.

4. Design the automation

Requirements

Define the technical and functional requirements for automating the processes.

Tool selection

Choose suitable software tools and technologies, such as RPA, workflow systems, IFTTT, Make, or Zapier.

5. Implement the automation

Development and configuration

Build and configure the automated solution according to the design. Use a separate development environment.

Testing

Test that automated processes work correctly and deliver the required results. Use a staging or pre-production environment.

6. Deploy and operate

Deployment

Deploy the automated solution to the production environment.

Training and change management

Train employees in the new automated processes and manage the changes to support a smooth transition.

7. Monitor and optimise

Monitoring

Track automated process performance using metrics, for example with Prometheus and Grafana, and KPIs.

Optimisation

Continuously improve automated processes based on feedback and performance analysis.

Summary

Choose a suitable process, describe its steps and exceptions, design a solution, and test it outside production. After deployment, assign operational ownership and evaluate benefits using measures agreed in advance.

Illustrative example: a team first validates order transfers between two applications using test data. The pilot covers duplicates and an outage of one application before the integration enters routine operation. This is not a case study of Erik’s work.